Lyles College of Engineering
2026 Projects Day
Lyles College of Engineering
Agricultural Technology Education and Laboratory Building
Students: Estefani Zermeno Romo, Evan Angel Candelaria, Cesar Alfonso Castaneda, Christopher
Kevin Espinoza Gonzalez, Elliott Stephan-Taylor Mills, Ennis Hasan Mohammad, Fatima
Plascencia Gutierrez, Julian Rosas
Mentors/Advisors: Loren Aiton, Dr. Yupeng Vivien Luo
Project Summary:
The senior capstone project involving a mixture of Architectural Studies and Construction
Management students highlights real-world partnerships and collaboration. As a team
focused on design build methodologies, our team is dedicated to delivering the capstone
project with the intention of serving and supporting agricultural innovation. With
a project budget of $122 million dollars and a required building square footage of
81,000, our team seeks to deliver spaces that inspire advancement in agriculture through
hands-on learning and collaboration within the industry. Along with taking measures
to consider the surrounding agriculture through attentive safety and sustainability
efforts. This project aims to increase California State University, Fresno’s students
and industry partners' preparation and innovation in the evolving world of agriculture.
Agriculture Technology Building Abstract - Orvis Tierra
Students: Anthony Melella, Cesar Rocha, Esmeralda Ramirez, Vicente Maldonado, Daniel Ramirez,
Jesus Aquino-Aguilar, Nolan Martinez
Mentor/Advisor: Dr. Yupeng Vivien Luo
Project Summary:
In this culminating capstone project, students from Architectural Studies and Construction
Management formed Orvis Tierra, an integrated design-build team responding to a formal
Request for Proposals (RFP) for a hypothetical Agriculture Technology building at
Fresno State. The team developed a comprehensive proposal, including design documents,
cost estimates, and construction schedules that reflect real-world practice. Working
within a $122 million budget and targeting LEED Gold certification, the team prioritized
sustainable strategies and ergonomic spatial planning to support collaboration among
students, faculty, and industry partners. By applying individually conducted research
on different building systems, each team member influenced the final design by integrating
early suggestions during the design phase. The members also worked collaboratively
by assuming roles typical of real-world design-build applications. By assuming professional
roles, the team influenced the building’s design and completion of associated deliverables,
setting accurate expectations for its members and the industry.
Application of Generative AI for Waste Reduction in Construction - A Systematic Review
Student: Sukumar Bachu
Advisor/Mentor: Dr. Tolulope Sanni
Project Summary:
The construction industry generates large amounts of material waste and greenhouse
gas emissions, making sustainable solutions increasingly important. This research
explores how Generative Artificial Intelligence (GenAI) can help reduce construction
waste and improve efficiency across the Architecture, Engineering, and Construction
(AEC) sector. A systematic literature review was conducted following PRISMA 2020
guidelines, analyzing 29 peer reviewed studies published between 2020 and 2025. The
study examines current applications of GenAI in waste reduction, identifies key benefits
and challenges, and highlights future research opportunities. Findings show that GenAI
supports waste prevention through optimized design, robotic sorting of demolition
materials, and improved lean construction workflows. While benefits include better
decision-making and material efficiency, challenges such as data limitations, high
costs, and reliability concerns remain. This study provides a foundation for advancing
AI-driven sustainability in construction.
Cultivating for the Future: Agricultural Technology Education and Laboratory Building
Students: Alondra Guerrero, Audriey Alpuerto, Breana Quintero, Zwe Htet Aung, Carlos Vargas,
Fernando Rodarte, Jovanny Rodriguez
Mentor/Advisor: Dr. Yupeng Vivien Luo
Project Summary:
This interdisciplinary senior capstone project, developed collaboratively by Construction
Management (CM) and Architectural Studies (AS) students, proposes a new Agricultural
Technology Education and Laboratory Building at California State University, Fresno.
Using a design-build approach, the project simulates integrated project delivery and
cross-disciplinary collaboration. Envisioned as a bridge connecting Lyles College
of Engineering and Jordan College of Agricultural Sciences and Technology, this building
includes robotics, automation, drone testing, and student support spaces to foster
collaboration between faculty and students. The project was developed to meet a specified
$122 million budget and an 81,312 sqft program. AS students led design development,
site analysis, programming, and sustainability strategies, while CM students focused
on cost estimating, scheduling, site logistics, and constructability review. The poster
highlights design intent through floor plans, renderings, sustainability features,
cost summary, and scheduling. It demonstrates how the design-build approach aligns
architectural vision with construction execution.
Foundation 7: Fresno State Agricultural Technology Building
Students: Rogelio Lopez Magana, Ramiro Moreno, Cristian Sosa, Nicole Figueroa, Fabian Sepulveda,Jacqueline
Lopez, Isabelle Estes
Mentor/Advisor: Molly Smith
Project Summary:
Foundation 7 is a multidisciplinary, student-led design-build team formed to simulate
the structure and demands of a professional construction project conducted over two
academic semesters. The team consists of seven students from the Architecture and
Construction Management programs, organized to reflect an integrated design-build
delivery model. The project centers on the Fresno State Agricultural Technology Building,
focusing on unique features like equipment-heavy research spaces, flexible lab areas,
and green spaces being incorporated into the building. Foundation 7 is developing
a proposal that integrates detailed quantity takeoffs, a milestone based construction
schedule and a sustainability focus analysis to constructability and performance.
Our team is addressing unique project challenges like coordinating MEP systems, balancing
innovation and sustainability with budget limits and coordinating long lead items
during the preconstruction phase. Project’s Day attendees will see renderings, 3D
models, scheduling visuals and cost summaries that will highlight the project's unique
features. Through this process, the team identified tradeoffs between cost, sustainability
goals, and constructability, leading to a deeper understanding of how complex project
decisions are made in a professional setting.
Hydrogen Fuel Cell Electric Vehicle Workforce Development in Public Transit
Student: Jaskaran Preet Singh
Mentor/Advisor: Dr. Manideep Tummalapudi
Project Summary:
Public transit agencies are increasingly adopting hydrogen fuel cell electric vehicles
(FCEVs) as a pathway toward zero-emission mobility systems. However, this technological
transition presents critical workforce development challenges requiring specialized
competencies beyond existing transit technologies. This systematic review publications
from 2010-2024 across academic databases and transit agency sources to identify essential
FCEV workforce competencies and compare them with predecessor technologies. The analysis
examined knowledge, skills, and abilities across vehicle operators, maintenance technicians,
and facility personnel. Findings reveal that FCEVs demand integration of mechanical,
electrical, and chemical safety competencies. Comparative analysis demonstrates asymmetric
transferability patterns: CNG backgrounds provide strong foundations for operators,
while BEB technicians contribute high-voltage electrical competencies. However, PEM
fuel cell maintenance, hydrogen material compatibility, and integrated safety protocols
represent entirely new competency domains. The study identifies strategic recruitment
pathways and modular training approaches that leverage transferable competencies,
emphasizing that workforce development must receive priority to technology deployment.
New Agricultural Technology and Education Building
Students: Deborah Garcia, Alex Banuelos-Valencia, Nathan Flores, Airelle Diaz, Shayne Stewart,
Matthew Briseno, Rogelio Huerta, Joslynn Archan
Mentors/Advisors: Molly Smith and Lloyd Crask
Project Summary:
G-Force is an interdisciplinary Design-Build capstone team composed of Architectural
Studies and Construction Management students responding to Fresno State’s RFP for
the New Agricultural Technology Education and Laboratory Building. As part of our
senior project, AS students developed the site analysis, schematic designs, 3D visualizations,
circulation strategies, and sustainable design solutions. We achieved LEED Gold by
implementing energy-efficient systems and green materials. CM students produced a
detailed Gantt schedule, cost estimate, constructability analysis, and site-specific
safety plan using industry-based tools and coordination methods. We fostered integrated
collaboration through weekly meetings and progress check-ins to evaluate design decisions
against the budget, schedule, and performance criteria, simulating a real-world design-build
delivery process. At Projects Day, we will present drawings, renderings, scheduling
documents, cost analyses, and sustainability strategies. Attendees will gain insight
into how interdisciplinary student teams apply academic research, technical tools,
and professional standards to deliver a comprehensive, high-performance building proposal.
New Agricultural Technology Education and Laboratory Building
Students: David Valdovinos-Valencia, Alexa Gamez, Bryan Gutierrez-Ruiz, Wilner Vargas, Carlos
Briceno, Kevyn Luna, Salvador Anaya, Braulio Moreno
Advisors/Mentors: Loren Aiton, Dr. Yupeng Vivien Luo
Project Summary:
The New Agricultural Technology Education and Laboratory Building is a facility that
goes beyond the boundaries of traditional agricultural practices. Our senior capstone
course for Architectural Studies and Construction Management majors entails collaborating
to develop this facility. Functioning as a Design-Build firm, our team works together
to produce architectural design solutions, estimates, schedules, and LEED sustainability
solutions. With the adjustment and rise of AI technology, through intensive, considerate
design, Pacific Legacy Builders has prioritized delivering the New Agricultural Technology
Education and Laboratory Building in the best manner through efficient construction
practices, while also creating a new modular experience on the Fresno State Campus.
Pacific Legacy Builder’s design philosophy is grounded in modern iterations of traditions
that bridge construction and design integrity.
New Agricultural Technology Education and Laboratory Building California State University,
Fresno
Students: Analiese Figueroa, Maya Abou Taha, Alejandro Barajas, Manuel Figueroa, Odalis Mata-Figueroa,
Arturo Hernandez, Yatziry Martinez, Flavio Ventura
Mentor/Advisor: Molly Smith
Project Summary:
Our year-long senior capstone project focused on developing a new 81,312-square-foot
Agricultural Technology Education Building at Fresno State, a collaborative hub where
agriculture and engineering converge to highlight the integration of science, technology,
and practice. Led by the Jordan College of Agricultural Sciences and Technology in
partnership with the Lyles College of Engineering, the project addresses California’s
agricultural challenges, including labor shortages, climate pressures, water scarcity,
rising costs, and new emerging technologies. The facility supports applied research
in robotics, drone systems, artificial intelligence, and controlled-environment agriculture,
with an emphasis on laboratory-based, hands-on instruction. Key challenges included
integrating specialized labs within a flexible framework, reducing the program from
97,600 to 81,312 square feet, achieving LEED Gold certification, and aligning scope
with budget and schedule under a design-build model. The project demonstrates strategic
planning and interdisciplinary coordination, applying skills developed in Fresno State’s
Construction Management and Architectural Studies program.
Potential Agricultural Technology Education and Laboratory Building
Students: Morgan Parent, Kayla Contreras, Sunem Serrano-Mendoza, Blazej Skochinski-Chavez, Isaiah
Soto, Alan Viruete-Millan, Daniel Zavala-Candelario, Cooper Lilles
Mentor/Advisor: Molly Smith
Project Summary:
Starting in August 2025, the senior capstone students in both Architectural Studies
and Construction Management have been brought together to test their abilities in
a project that will simulate what type of work environment is expected after Fresno
State. The senior capstone class tests our abilities as students to conduct research
towards our hypothetical building as if we were working on its construction from day
one. Various research viewpoints were implemented, such as site analysis, city-centric
code analysis, scheduling, and estimating. As a team, the Construction Management
and Architectural Studies students work together to bring two sides of the industry
together for a project that will test everyone’s abilities. The project, a new Agricultural
Technology Center, will be displayed to showcase the work that each student has accumulated
during their time at Fresno State to prove their readiness for the field.
A Comparative Study of Unreal Engine and Unity for 3D Gaussian Splat–Based VR
Student: Daniel Ornelas
Mentor/Advisor: Dr. Xiangxiong Kong
Project Summary:
Civil infrastructure inspection often depends on manual visual assessment, which is
costly, time-intensive, and vulnerable to human error. This study explores the use
of 3D Gaussian Splats (3DGS) - a mesh-free, photorealistic representation generated
from multi view imagery - to create Virtual Reality (VR) environments that support
more efficient digital inspections. A 3DGS model was integrated into Unreal Engine
5 (UE5) and Unity to evaluate their suitability for deployment on the Meta Quest 3.
UE5 provided built-in VR templates, but required significant adjustments during export.
Unity offered a smoother export pipeline, but demanded more custom coding to implement
interaction and annotation tools. The comparison highlights trade-offs in workflow
complexity, rendering performance, and development flexibility. We also report the
challenges we faced during the VR modeling workflow in this poster presentation.
A Comprehensive Water Demand Analysis of California using County-level Demand and
External Determinants
Students: Aaron Fregoso, Makayla Cawley
Mentor/Advisor: Dr. Jorge Pesantez
Project Summary:
California has faced extreme weather events that led to desperate management decisions
about the usage of available clean water. These climate change conditions, exacerbated
by an increasing population, require more sustainable management to prevent disasters
occurring over one of the most vital requirements for life. This project presents
a comprehensive analysis of the change in demand over time and the influence of weather
and socioeconomic conditions at multiple locations across California. We have included
monthly water demand reported by multiple counties from 2013 to 2021 to analyze the
consumption trends and variance in water demand and the effects of those important
exogenous determinants. The data for water demand for these assumptions comes from
the CaRDS dataset [Gross et. al]. These values were then compared with the respective
weather information, including: max, min, average temperature, precipitation, dew,
humidity, wind gust, and wind speed; as well as county-level socioeconomic information
consisting of education levels, unemployment rate, and mean and median incomes. Our
modeling was implemented in the MATLAB programming language and includes multiple
visuals and trends throughout our datasheet of over 4000 data points. Results showed
strong proportionality between water demand and population density, seasonal peaks
aligned with climatic cycles, and associations with socioeconomic patterns. These
findings put together the demographic, environmental, and economic drivers of water
use. Future work should focus on utilizing predictive modeling to integrate policy
interventions and exploring strategies for equitable water management.
A Framework for Evaluating Photogrammetry Workflows Using Image Similarity Metrics
Student: Felipe Ramirez
Mentor/Advisor: Dr. Xiangxiong Kong
Project Summary:
Photogrammetry is the science of extracting geometric and color information from a
series of images. The objective of this project was to develop a framework for evaluating
structural differences and visual distortions introduced by contemporary photogrammetry
workflows. To achieve this, RealityScan was used to generate a textured 3D model from
an image dataset. The resulting model was aligned with the original source images
and rendered in Blender. MATLAB was then employed to compute the Structural Similarity
Index Measure (SSIM), Multi-Scale Structural Similarity Index Measure (MS-SSIM), and
Peak Signal-to-Noise Ratio (PSNR), which quantify the structural fidelity and visual
quality of the reconstructed model relative to the original images. A case study of
a building structure was reported in this poster to test out the performance of the
proposed framework.
Assessment of Existing Leak Localization Efforts in Water Distribution Systems
Student: Carly Boyer
Mentor/Advisor: Dr. Jorge Pesantez
Project Summary:
Water distribution systems (WDS) are essential infrastructure that are vulnerable
to leaks occurring at multiple locations within the network. Leak detection using
pressure sensors has been extensively studied in the literature. However, pressure
measurements for leak localization offer significant opportunities to narrow down
the search area. Leak localization is a complex task that depends on network topology
and hydraulic properties. Factors such as sensor placement, spacing, and measurement
temporal resolution can be used to evaluate a supervised learning algorithm for leak
localization. This research presents a systematic approach to assessing existing leak
localization methods. The approach uses multiple networks and hydraulic and topological
characteristics from the literature to summarize the effects on localizing leaks of
different components, such as network size, range of nodal elevation, fluid velocity,
and the presence of tanks and controls. The results will provide researchers and practitioners
with an objective comparative analysis that can be adapted to specific network conditions.
Furthermore, this project supports the WDS community in locating leaks and reducing
their negative impacts, such as property damage, service interruptions, water loss,
intrusion contamination, and revenue loss for utility providers.
Examining the Correlation Between Resilient Modulus (Mᵣ) and California Bearing Ratio
(CBR) for Central Valley Soils
Student: Soushilya Maggidi
Mentor/Advisor: Dr. Lalita Oka
Project Summary:
Subgrade soil properties are fundamental to pavement durability, as inadequate characterization
often leads to premature failure. While the California Bearing Ratio (CBR) has traditionally
been used to evaluate subgrade strength due to its cost-effectiveness, it fails to
capture the soil’s elastic response to the repeated dynamic loading typical of traffic.
Consequently, modern Mechanistic-Empirical (M-E) design has shifted toward using the
Resilient Modulus (Mr). However, because direct Mr testing is expensive and requires
specialized equipment, engineers often rely on empirical correlations to estimate
it from CBR values. A critical gap exists as current national correlations often fail
to account for the unique characteristics of California’s Central Valley soils, such
as high fines content and extreme moisture sensitivity. This study addresses this
discrepancy by developing region-specific CBR–Mr correlations through laboratory testing
of representative soil samples from the Fresno and Clovis areas, including sandy loams
and silty clays. By establishing localized mathematical relationships and comparing
them against national models, this research provides more reliable design inputs that
reflect local conditions, ultimately improving the performance and longevity of regional
pavement infrastructure.
Investigating the Impact of Recycled Asphalt Pavement (RAP) on Cracking Trends in
Asphalt Sections: A Data-Driven Study Using LTPP
Student: Rahul Akula
Mentor/Advisor: Dr. Xiaojun Li
Project Summary:
The use of Recycled Asphalt Pavement (RAP) has increased in recent years because it
helps reduce construction costs and supports environmental sustainability. By reusing
old asphalt materials, agencies can save natural resources and lower project expenses.
However, there are still questions about how RAP affects long-term pavement performance,
especially when it comes to cracking. Since cracking is one of the most common types
of pavement damage, it is important to understand whether adding RAP makes pavements
perform better, worse, or about the same over time. This study examines the effect
of RAP on cracking trends in asphalt pavement sections using data from the Long-Term
Pavement Performance (LTPP) database. The research compares pavement sections that
contain RAP with those that do not. Cracking behavior is evaluated by looking at
cracking type and overall cracking levels across several years of field surveys. The
study also considers traffic levels and climate conditions, since both factors can
strongly influence pavement performance. A data-based analysis is carried out using
simple statistical measures, such as average cracking values, to compare performance
between RAP and non-RAP sections. The results help identify performance patterns
and possible regional differences.
Investigating The Relationship Between Subsidence And Drawdown In Piezometric Levels
In The Central Valley
Student: Veronica Molina Rodriguez
Mentor/Advisor: Dr. Lalita Oka
Project Summary:
Land subsidence associated with groundwater extraction represents a cumulative and
often irreversible process that threatens infrastructure integrity and long-term aquifer
sustainability. In California’s Central Valley, subsidence is closely linked to piezometric
drawdown in compressible alluvial systems, where declining pore pressure increases
effective stress and induces large consolidation settlement (subsidence). This study
evaluates the relationship between subsidence and groundwater level decline across
selected counties in the Central Valley (Contra Costa, Sacramento, San Joaquin, Solano,
Stanislaus, and Yolo) through an integrated framework combining observational data,
GIS-based spatial analysis, and transient numerical modeling. Publicly available USGS
data were compiled to construct a reproducible regional database, incorporating subsidence
records since 2011 and piezometric data for 2020 - 2025. Spatial interpolation identified
priority areas characterized by cumulative subsidence of up to 0.51 ft and marked
groundwater depletion. A 12-month MODFLOW (PMWIN) simulation for 2024, including seasonal
recharge and pumping, reported a minimal water budget discrepancy (0.01%) and a sustained
annual storage deficit, with declines reaching 68% in some areas. The results thus
reinforce the physical linkage between drawdown and subsidence along with quantifying
the shortage that caused the overdraft. Thus, this modeling approach offers a strategic
water management metric to support preventive, evidence-based groundwater management
in drought-prone regions.
Mining Claim Accessibility Through GIS
Student: Katlin Rowley
Mentor/Advisor: Dr. Mike Berber
Project Summary:
The information contained in mining claim documents is highly valuable to the mining
surveying industry. Although the information contained within these documents is
crucial, that information is rarely easily accessible. In general, in order to acquire
recorded mining claim documents, an individual in California would have to work with
their County or the Bureau of Land Management to acquire those documents. This can
be an expensive and time consuming process. The enclosed project aims to solve this
issue and increase the accessibility of mining claim documents. This project completed
this objective by prototyping an interactive GIS map for Trinity County that allows
you to select the mining claim of interest and directly access the recorded mining
claim. The project was then expanded to include research into the current open source
mining claim GIS data from government adjacencies. These layers are then compared
to the GIS prototype created within this project. Overall, demonstrating the ability
of GIS to increase transparency and accessibility within the mining surveying industry.
Mining Surveying: Methods and Standards
Student: Katlin Rowley
Mentor/Advisor: Dr. Mike Berber
Project Summary:
Geomatics engineering is a complex and diverse field with many subspecialties; two
such subspecialities being adjustment computations and mining surveying. The overlap
of these subspecialities will be the focus of this research proposal. Within mining
surveying, a licensed land surveyor or engineer is not typically required and a set
of accuracy standards is not widely adopted within the United States. With this in
mind, this research aims to answer the question of whether current mining surveying
techniques and adjustment computation procedures (accuracy standards) when marking
mining claims effectively protects ownership rights.
Ownership rights are at the center of a multitude of geomatics engineering work, and many organizations have established standards to help ensure that data accuracy is sufficient to protect these rights. Some examples of this being the American Land Title Association (ALTA) standards and the American Society for Photogrammetry and Remote Sensing (ASPRS) standards. Standards such as those from ALTA and ASPRS address the importance of accuracy within their subspecialties of geomatics engineering. Determinations made based upon these standards can and do effect rights, and this is one of the many reasons why a standard for the subspecialties were created. A set of standards to follow helps to ensure that the work being done throughout the subspeciality is to the same level of accuracy and quality, and therefore the same level of protection of rights.
The absence of an adopted method to establish ownership rights, within Mining Surveying, may create a risk for the public/owners. Individuals creating mining claims may not be using surveying techniques and adjustment computation procedures to a sufficient standard to protect the rights that a mining claim would grant individuals if produced to a set of adopted standards. Widely adopted mining procedures and standards may need to be put into place to properly protect the ownership rights a mining claim creates. The proposed research will work to identify whether current mining surveying techniques and accuracies are sufficient to protect ownership rights, or if a mining surveying method(s) and standards publication is necessary.
Researching and compiling historical records of the evolution of mining surveying, current mining surveying procedures/techniques, mining regulations and mining laws at both the federal and state of California level will be fundamental to making this determination. It is expected that this research will indicate that a mining surveying methods and standards publication is necessary. The goal of such publication being to protect ownership rights. If such determination is made, this research will encompass recommendations as to what the adopted methods and standards for mining surveying should be. This will be accomplished by analyzing current land surveying techniques and technologies along with widely adopted accuracy standards in adjacent geomatics engineering subspecialties. This analysis will be applied to the findings of why a mining surveying methods and standards publication is necessary. Methods and standards will be recommended based upon how current surveying technology, techniques, and accuracy standards can effectively address the identified deficiencies in current mining surveying procedures.
Modeling Pressure and Flow Deviations for Leak Detection in Water Distribution Networks
Student: Angela Maldonado
Mentor/Advisor: Dr. Jorge Pesantez
Project Summary:
Detecting and localizing leaks in water distribution systems presents a persistent
challenge for water utilities due to aging infrastructure. Undetected leaks can lead
to significant potable water losses, reduced pressures, and contamination into the
distribution system. Pressure sensors are commonly used to locate leaks and reduce
water losses. However, limited physical and economic feasibility may prevent utilities
from utilizing sensors for leak detection. This research proposes an adaptable water
distribution network model to simulate network behavior and detect leaks using pressure
and flow deviations. The hydraulics-based model will identify whether a pipe deviates
from its expected pressure and flow. By analyzing the deviations from expected conditions,
the model will estimate probable leak locations. The methodology intends to cross-validate
the results of the distribution network model with an operational system containing
pressure sensors. The proposed model offers utilities a scalable, cost-effective tool
to optimize system performance and improve water distribution management.
Pedestrian Network Performance Analysis of Reduced Vertical Egress Redundancy in Multi-Story
Buildings
Student: Artur Martirosyan
Mentor/Advisor: Dr. Julio Roa
Project Summary:
What impact does reducing vertical egress redundancy have on multi-level building
evacuation performance? In order to increase life safety through redundancy, traditional
building codes mandate that residential buildings taller than three stories have two
exit stairways. However, the transportation performance implications of removing one
of these vertical routes remain under-explored. In this project, a multi-story building
is treated as a network of pedestrian transportation, with occupants serving as dynamic
flow units and stairways acting as vertical links. The study compares two-stair and
single-stair configurations to assess congestion, bottleneck formation, flow rates,
and overall evacuation time using evacuation modelling principles and transportation
flow theory. System resilience and efficiency under varied occupant loads are being
measured using capacity analysis. Furthermore, to evaluate trade-offs between infrastructure
efficiency and redundant egress capacity, a cost-performance comparison is also carried
out in this project. Ultimately the results are used to demonstrate the effects on
network performance in a real-world setting using a case-study building model.
System-Level Transfer Learning from Public Water Demand Datasets
Student: Alessandro Toledo Salazar
Mentor/Advisor: Dr. Jorge Pesantez
Project Summary:
The limited access to demand datasets has constrained the analysis of water distribution
systems (WDSs). Water agencies may face challenges at implementing uniform metering
systems and at sharing the entire dataset due to confidentiality for their customers.
For that reason, only public or synthetic datasets have been used to develop data-based
models that predict hydraulic parameters along open source networks. All these models
rely on high-quality, complete source data to train and generate predictions. However,
these predictive models also have the capability to transfer knowledge from observed
patterns. This characteristic, known as transfer learning, may support the use of
publicly available datasets to predict demand related to networks with low-quality
or scarce data. This study is focused on using a transfer learning approach to train
predictive models from public data for system-level demand. The data processing and
training involve a network similarity analysis and classification based on consumption
levels and intraday patterns. The results aim to demonstrate that pretraining neural
networks improves the prediction for similar water networks, which have limited data.
The development of this approach intends to provide water agencies with the ability
of understanding their networks from existing models and accessible datasets. Likewise,
it is supposed to simplify the analysis of networks that have undergone changes over
time.
Triaxial Testing of Soil–Rubber Mixtures for Sustainable Seismic Geotechnical Design
Students: Inam Ullah, Soushilya Maggidi
Mentor/Advisor: Dr. Arezoo Sadrinezhad
Project Summary:
The use of recycled materials in geotechnical engineering has gained increasing attention
due to sustainability concerns and the need for resilient infrastructure in seismic
regions. Tire-derived aggregates and soil–rubber mixtures offer potential advantages
such as reduced unit weight, enhanced energy dissipation, and improved deformation
characteristics, making them attractive for applications including retaining wall
backfill, embankments, and foundation systems. This study employs monotonic and cyclic
triaxial testing to investigate the mechanical and seismic behavior of sand-rubber
mixtures under loading conditions relevant to earthquake demands. Specimens containing
varying TDA contents are tested under controlled confining pressures to examine stress-strain
behavior, stiffness degradation, deformation response, and energy dissipation under
repeated loading. The findings of this research will establish experimentally based
trends that clarify the role of rubber content in governing seismic response and will
provide critical data to support the safe and effective use of recycled tire materials
in sustainable geotechnical design.
10s3p Battery Pack with Passive Balancing BMS
Students: Javier Jimenez, Roberto Carlos Barajas, Jesus A. Barajas Luque
Mentor/Advisor: Dr. Woonki Na
Project Summary:
This project develops a modular Battery Management System (BMS) for a 10s3p lithium-ion
battery pack intended for light electric vehicle applications. Lithium-ion batteries
offer high energy density but require careful monitoring to prevent safety hazards
such as overcharging, deep discharge, and overheating. Our system continuously monitors
the voltage and temperature of each cell group and uses passive balancing to equalize
charge differences by safely dissipating excess energy as heat. A distributed architecture
is implemented, where individual ATtiny841-based cell modules manage local sensing
and balancing, while a central ESP32 controller aggregates data and hosts a WiFi-enabled
web dashboard for real-time monitoring. The design emphasizes cost-effectiveness,
safety, and reliability while meeting industry standards for lithium-ion protection.
Accessible Soil Moisture Sensor System for Small Farmers
Students: Cisty Vue, Ranbir Mann
Mentor/Advisor: Dr. Soumyasanta Laha
Project Summary:
The purpose of the project is to supply small farmers with easy-to-use soil sensor
system to help them manage their water resources. The system will also allow farmers
to remotely monitor their crop so that they will only need to be there if work is
really needed. The project is based on a small survey conducted with 36 small farmers
to gain insight into some of their challenges. The survey results show that there
are challenges in using these technologies from those not so "tech-savvy." The survey
also shows that there is a need for help managing water resources as well as simple
and easy to use and understand technology to be more accessible for some of the farmers
who are not so tech-savvy.
Agriculture-Based Renewable Energy Source via Portable Photovoltaic Cells and Digital
Twin Applications
Students: Bailey Wooten, Johnnie Cerda, Mac Leal
Mentor/Advisor: Dr. Woonki Na
Project Summary:
In recent years, solar has been growing as a means of generating electricity to power
homes, cars, and many other things. Solar is still not as popular in agriculture as
many would think. If solar were the main source of energy for the agriculture industry,
as other industries, then the amount of renewable energy would increase greatly. This
project will take this idea head-on and apply smart renewable energy to irrigation,
which is one of the key parts of agriculture. This will be accomplished by integrating
a ESP-32 micro-controller to an irrigation system that runs on solar acting as a battery
management system (BMS). Information from the micro-controller will then be sent to
the cloud using Internet of Things (IoT) to allow farmers to monitor the data. The
physical implementation will be verified using a hardware-in-the-loop (HIL) software
to safely test the micro-controller to ensure proper operation before implementation
in the field.
Autonomous Enclosed-Rotor UAV for Target Detection and AR Visualization in High-Risk
Environments
Students: Nicholas Amely, Alexander Moreno Gonzalez, Eliasim Fragoso
Mentor/Advisor: Dr. Gregory Kriehn
Project Summary:
This project develops an autonomous indoor drone with protective enclosed rotors and
augmented reality (AR) visualization for search-and-rescue and tactical operations
in hazardous environments. The drone uses an RGB camera to detect people and objects
behind walls, navigating autonomously without GPS by using computer vision and mapping
technology. Real-time video and sensor data stream wirelessly to a Meta Quest 3 VR
headset, where custom software displays a 3D reconstructed environment. An integrated
artificial intelligence system identifies and highlights humans, animals, and objects
of interest within the AR view. Built for under $1,000, the system enables operators
to safely explore dangerous or unstable buildings remotely. The three-phase development
includes drone construction, autonomous flight integration using a Raspberry Pi computer,
and intelligent target detection capabilities.
Autonomous Turret with AI Object Tracking
Students: Jose Bravo, Fernando Solorzano, John Garcia
Mentor/Advisor: Dr. Hovannes Kulhandjian
Project Summary:
This project is a fully autonomous turret system designed to identify, track, and
target drones. The turret system consists of two integrated subsystems: the targeting
software and the mechanism hardware. The software is built using the YOLOv8 object
detection model, acquiring data and code from a repository; the object detection system
is also trained by supplying image data to the model’s database. The hardware is a
two degree-of-freedom pan and tilt system that maneuvers the camera and armament using
a gear and pulley belt drive system. The frame and body of the turret system was 3D
printed using edited prefabbed models. The turret system is run by a Raspberry Pi
5, which handles the computational processing of the software and controls the stepper
drivers to operate the belt drive systems.
Computer-Upgraded Mouse
Students: David Medina, Alejandro Zepeda, John Herrera
Mentor/Advisor: Dr. Nan Wang
Project Summary:
The Micromouse competition, founded by IEEE in the late 1970s, is an international
robotics challenge that requires teams to design and program autonomous robots that
explore and solve an unknown maze, reaching the center in the shortest possible time.
Our team will develop an original micromouse, the “Computer-Upgraded Mouse,” competing
under Fresno State’s IEEE chapter. We plan to use a set of infrared sensors to detect
walls and use a flood fill algorithm to be able to autonomously navigate terrain,
with potential scalability for real world terrain navigating applications. The use
of computer vision, in conjunction with solving algorithms will allow for the most
optimal path being selected when solving the maze.
Embedded Vision Prosthetic Hand
Students: Anthony Nichols Jr., Fernando Hernandez, Manjot Dhanda
Mentor/Advisor: Dr. Woonki Na
Project Summary:
This project is a wearable prosthetic hand that uses a small onboard computer and
a camera to help choose how to grasp everyday objects. The system runs a lightweight
YOLO object detection model on a Raspberry Pi 5. Instead of only identifying items,
the program groups them into handling types, such as a gentle pinch for fragile objects
or a full wrap grip for larger ones. To avoid sudden movements, the software checks
detections across multiple frames and requires consistent confidence before the hand
reacts. An Android app allows the user to confirm actions, switch modes, or manually
control the hand when needed, while the prosthetic continues basic operation on its
own. The goal is to reduce the effort required to operate a prosthetic by letting
the device assist with grip decisions, demonstrating a practical approach to assistive
robotics using low-cost hardware.
Gesture Controlled FPV Drone
Students: Juan Aragon, Jocelyne Solano, Logan Bradshaw
Mentor/Advisor: Roger Moore
Project Summary:
Recently, there has been an increase of FPV drones used for various different tasks
including search & rescue, geographical mapping, precision crop monitoring, and commercial
videography. FPV drone technology has allowed people to obtain footage of locations
inaccessible to humans. In this project, a drone will be capable of being controlled
in two different ways: gesture control and remote control. By utilizing gyroscopic
controls, the user has more control over the movement of the drone. It is also more
intuitive than the standard control scheme for drones, allowing more widespread usage.
In addition, the drone can automate its flight by utilizing user-set waypoint missions
that determine where the drone will depart and land. The drone will use a Pixhawk
flight control system, and the controller will use a Pi Pico 2W to send control signals
to the drone based on which gesture occurs.
Industrial Building Management System
Students: Tue Do, Daniel Perez Castellanos
Mentor/Advisor: Roger Moore
Project Summary:
This project presents the design and implementation of a portable Industrial Building
Management System (IBMS) that demonstrates how modern buildings can be monitored and
controlled efficiently using low-cost embedded technology. The system simulates an
industrial environment with four independent rooms, each equipped with sensors to
track temperature, air quality, lighting, and occupancy. A centralized controller
processes this data and displays it through a user-friendly interface, allowing real-time
monitoring and control of lighting, ventilation, alarms, and safety features. By integrating
distributed sensors, automated control logic, and centralized visualization, the project
highlights how intelligent building systems can improve energy efficiency, occupant
comfort, and safety. Although developed as an academic prototype, the system reflects
real industrial architectures and serves as a scalable, hands-on example of modern
building automation principles.
MetavoltVR-AI: An AI Enhanced Student Developed VR Circuit Building Application
Students: Nicholas Amely, Jesus Leyva
Mentor/Advisor: Dr. Wei Wu
Project Summary:
Engineering is facing a critical workforce shortage while students are struggling
with traditional lecture-based instruction in foundational courses. MetavoltVR-AI
addresses these challenges by combining Virtual Reality (VR) with Artificial Intelligence
(AI) to create personalized, immersive learning experiences for engineering students.
Building on a successful VR prototype developed by the research team, this project
enhances the platform with two AI capabilities: a rule-based adaptive system that
tracks student performance and adjusts difficulty in real-time, and a natural language
processing interface that provides instant answers to student questions during VR
sessions. Targeting first-year engineering students, MetavoltVR-AI aims to increase
engagement, improve conceptual understanding of logic gates and circuits, and support
underrepresented students who face barriers in traditional engineering education.
This project demonstrates how emerging technologies can make engineering education
more accessible, engaging, and effective for diverse learners.
Mini Power Grid
Students: Jose Placensia, Arian Rostami, Alejandro Alvarado
Mentor/Advisor: Roger Moore and Dr. Woonki Na
Project Summary:
This project focuses on the design and implementation of a low-voltage solar-powered
DC microgrid for small-scale distributed energy applications. The system operates
on a nominal 40–50 V DC bus and integrates a photovoltaic (PV) source, a DC–DC converter
with maximum power point tracking (MPPT), regulated DC outputs, and a low-voltage
single-phase micro-inverter for AC loads. An STM32 microcontroller is used for real-time
voltage and current sensing, PWM generation, and closed-loop control of the power
converters. The MPPT algorithm dynamically adjusts the converter duty cycle to maximize
power extraction under varying irradiance conditions. Simulation and experimental
testing are performed to evaluate voltage regulation, efficiency, and system stability.
The resulting design demonstrates a modular and scalable microgrid architecture suitable
for educational and low-power renewable energy applications.
Non-Invasive Blood Glucose Monitoring
Students: Nick Pierce, Jacob Lira, Rafael Garcia-Gallardo
Mentor/Advisor: Dr. Soumyasanta Laha
Project Summary:
Diabetes management relies heavily on regular monitoring of blood glucose levels,
traditionally performed through invasive finger-prick methods. While accurate, these
techniques can cause discomfort, risk of infection, and reduced patient compliance
over time. To address these limitations, recent research has focused on developing
non-invasive glucose monitoring systems that can estimate glucose concentrations without
the need for blood extraction. Non-invasive approaches typically use electromagnetic,
optical, or acoustic spectrometry to detect physiological changes correlated with
glucose levels. This project aims to design and evaluate a hybrid non-invasive glucose
monitoring system that integrates electromagnetic and optical sensing, with plans
to later incorporate acoustic spectrometry. The electromag-netic subsystem will analyze
variations in blood permittivity and conductivity, while the optical subsystem will
measure light absorption in the near-infrared region. Data from these systems will
be processed using advanced machine learning algorithms, including Linear Regression,
Random Forests, Support Vector Machines, and Convolutional Neural Networks. The goal
is to determine the most accurate and efficient sensing modality for implementation
in a compact, wearable, and energy-efficient device to improve comfort and accessibility
for diabetic patients.
PCB Assembly Line
Students: Ashmeet Singh, Sydney Rivera
Mentor/Advisor: Roger Moore
Project Summary:
This project recreates a small-scale printed circuit board (PCB) manufacturing assembly
line, integrating automated quality inspection and component assembly. The system
consists of two primary modules: a machine vision–based quality inspection module
that uses machine learning to verify PCB trace accuracy, and a pick-and-place module
that positions surface-mount device (SMD) components onto the board. A programmable
logic controller (PLC) is designed from scratch using OpenPLC to coordinate system
operation, sensor integration, and actuator control. Camera systems are trained to
detect PCB faults and ensure accurate component placement. The system emphasizes real-world
industrial automation practices, including sensor integration, closed-loop control,
and modular system design to improve reliability and scalability.
Regenerative Braking Electric Bicycle
Students: Jose Sahagun, Laith Khoury, Kai Cayetano
Mentor/Advisor: Dr. Woonki Na, Roger Moore
Project Summary:
This project focuses on the design and implementation of a regenerative braking system
for a 48 V, 1000 W electric bicycle to improve energy efficiency and sustainability.
The system recovers kinetic energy lost during braking and pedaling and converts it
into electrical energy that is stored in the battery via a supporting capacitor bank.
The design integrates a rear-hub motor, bidirectional DC–DC buck/boost converter,
bidirectional DC–AC inverter, capacitor bank, and an ESP32-based control system. Custom
power-electronics hardware and printed circuit boards are designed, built, and tested
to safely manage power flow during both propulsion and braking. System performance
is evaluated through measurements of voltage, current, power, and charging efficiency.
This project provides hands-on experience with power electronics and control systems
while demonstrating practical applications of regenerative energy recovery in electric
transportation.
Smart Energy Router
Students: Jose Chavez, Andres Cardenas, Anirudh Nishtala
Mentor/Advisor: Roger Moore
Project Summary:
The Smart Energy Router is designed to actively monitor and manage power distribution
across residential energy nodes, including solar arrays, the utility grid, and battery
storage. This system integrates a custom designed 12V-to-20V DC-DC boost converter
and sensing circuits to optimize energy allocation, all of which is intended to be
programmed with python on a Raspberry Pi. By prioritizing energy consumption during
off-peak hours and utilizing stored reserves, the router reduces consumer utility
costs and eases demand on public utilities. The system is engineered to manage a 400W
household load and 50W solar input while functioning as a zero export system to comply
with safety and facility regulations. It demonstrates a cost-effective, scalable solution
for enhancing residential energy independence through hardware and software integration.
Wearable Blood Pressure Monitor Device
Students: Jacob Arias, Anuk Amarasinghe, Upek Amarasinghe
Mentor/Advisor: Dr. Soumyasanta Laha
Project Summary:
Our project is a wearable blood pressure device that will be able to display the required
information onto an app on your smart phone with bluetooth. We are trying to make
our product more affordable and also more accurate with its measurement for our users.
To achieve this we will be utilizing the ESP32-S3 micro controller integrated with
a high precision SparkFun pulse oximeter and heart rate sensor to collect our real
time data. Our targeted users will be anyone that will be wanting to check on their
heart effortlessly throughout the day so they can easily integrate it into their daily
routine.
Agrivolatics
Students: Cesar Aranibar, Alexa Alvarado, Angel Gama, Harman Kaur, Lucio Rangel
Mentor/Advisors: Dr. Yuanyuan Xie, Dr. Woonki Na, Dr. Qun Sun
Project Summary:
Agrivoltaics is a relatively new and innovative approach that integrates photovoltaic
(PV) energy systems with agricultural land use. The core concept involves installing
solar panels above actively farmed areas, whether crop fields or grazing lands, so
that solar energy can be harvested without significantly disrupting agricultural operations.
This dual-use system aims to maximize land efficiency by simultaneously supporting
both food and energy production. As the global demand for clean energy increases and
the availability of arable land becomes more limited, agrivoltaics presents a promising
strategy for sustainable development. This project focuses on incorporating our solar
technology into an agricultural environment and developing a dual-use system that
improves both crop productivity and renewable energy generation. By evaluating shading
levels, soil moisture retention, and panel output under various field conditions,
the system is engineered to maximize photovoltaic performance. The system incorporates
sensors and monitoring technology to track panel efficiency, record solar and wind
conditions, and support analysis of their effects on crop production.
AIAA and IEEE Airbrakes Project
Students: Corbin Obata, Carlos Cortes, Niqui Empis, Jonah Rayne, Jennifer Rosales, Alondra Perez-Ramirez
Mentor/Advisor: Dr. Deify Law
Project Summary:
The AIAA Fresno and IEEE Fresno Airbrakes project aims to create an electronically-controlled
airbrake system for rocketry to hit a target altitude. The mechanical design uses
a 7-link system to deploy and retract the braking system. Extension of the brakes’
pedals induce drag within the rocket by increasing the rocket’s cross-sectional area
during ascent. The increase of drag allows us to slow and stabilize the rocket to
reach the target apogee (maximum altitude of the rocket). For electronics, this project
features an integrated flight computer centered on the Raspberry Pi Pico 2 W. Utilizing
MicroPython, the system executes real-time sensor fusion and PID logic by processing
data from a BMP390 altimeter and MPU-6050 IMU. This firmware calculates predicted
apogee to modulate a high-torque MG996R servo. Power is supplied by a Tattu 2S LiPo
and Micro BEC, ensuring stable 5V delivery for precise actuation and robust data logging
during high-velocity flight.
Compact Autonomous Robot for Car Trunk Entry/Exit
Students: Alonso Camarillo, T.J. Fuentes, Gabriel Gamez Martinez, Simardeep Sahota, Gary West
Mentor/Advisor: Dr. The Nguyen
Project Summary:
Transporting heavy items to or from a vehicle can be physically demanding and unsafe,
especially for people with limited mobility. The objective of this project was to
design a compact robotic system capable of safely entering and exiting a car's trunk
after assisting with baggage. The system was required to meet size, budget, and weight
constraints, while maintaining safety and reliability. Designing involved defining
customer needs, reviewing prior art, establishing functions and subsystems. The final
design integrated a tracked mobility base, and a suction-based, articulated arm mount.
Safety considerations were applied through engineering standards, including ISO and
IEC. CAD models were developed to verify the system packaging, feasibility of constraints,
and articulation. The concept demonstrated its viability, providing the foundation
for future design and compliance with real-world deployment. Up to Projects Day, a
prototype that can move to and from the trunk will be complete, demonstrated on a
sedan.
Development of Enhanced UAV-based Drilling System Using Downward Thrust Generation
Students: Aaron Millwee, Brianna Vidrio, Rayne Saucedo, Ethan Le, Dana Martinez, Austin Duffy,
Jake Velicescu
Mentor/Advisor: Dr. Alaeddin Bani Milhim
Project Summary:
This study presents the design and experimental validation of a quadrotor-based drilling
platform that leverages the use of reversed propeller thrust to enhance soil drilling
effectiveness. The system employs a single low-speed DC motor (30 rpm) to generate
auger rotation, while reversed motor polarity is used to produce controlled downward
thrust that supplements the drone’s weight and existing actuation. The prototype uses
open-source software, commercially available components, and 3D-printed parts. Experiments
were conducted indoors in silty sand–topsoil mixtures with controlled strength levels
(0.5 kg/cm2 to 3.5 kg/cm2 ) and moisture contents of 35 % to 95 %, drilling to depths
of 5 cm to 15 cm. Across three applied thrust levels, 0.75 kg, 1.13 kg, 2.27 kg, and
a no-thrust condition, the trials demonstrate that drilling speed increases with applied
thrust and decreases with soil strength. Experimental results demonstrate that the
addition of reversed thrust increases the effective downward force, leading to improved
drilling effectiveness. A three-dimensional analysis of drilling speed as a function
of thrust and soil strength confirms that the added downward thrust effectively compensates
for increased soil resistance, leading to higher penetration efficiency. These results
validate the feasibility of thrust-reversal-assisted drilling and establish a foundation
for future UAV drilling platforms incorporating autonomous thrust control and optimized
auger geometries for operation in stronger and more heterogeneous substrates.
Experimental Testing and Performance Analysis for ThermeShade
Students: Bryan Mejia-Sanchez, Jose Ramblas, Mitchel Hamilton, Anahuic Reyes, Jonathan Camarena
Mentor/Advisor: Dr. Yuanyuan Xie
Project Summary:
This project develops and validates a testing framework to evaluate the real-world
performance of ThermeShade, a market-ready window shade designed to reduce solar heat
gain while preserving outdoor visibility and natural light. The testing approach focuses
on key performance metrics relevant to commercial applications, including room air
and surface temperatures, humidity, light transmission, HVAC energy usage, installation
feasibility, and product durability. Controlled experiments are to be conducted in
adjacent, south-facing on-campus rooms to closely replicate realistic operating conditions
and isolate the effects of the shade. Data is collected using calibrated thermal,
humidity, and energy-monitoring instruments and analyzed quantitatively and qualitatively.
Rather than redesigning the product, this work delivers a repeatable, data-driven
testing protocol and performance assessment to verify manufacturer claims, identify
areas for improvement, and support informed decision-making for future commercial
deployment of ThermeShade.
Experimentation and Simulation of Near-Infrared Light Penetration in Models of Oral
Tumors for Near-Infrared Photoimmunotherapy
Students: Samantha Schilling, Karen Martinez
Mentor/Advisor: Dr. Gemunu Happawana
Project Summary:
For over half a century, cancer treatment has primarily consisted of surgery, chemotherapy,
and radiotherapy. While semi-effective, they all possess serious side effects, weakening
the body’s natural immune system and damaging nearby cells. Due to this, a major focus
of cancer researchers has been uncovering cures to cancer that cause less harm to
the body. One recent discovery is the use of near-infrared photoimmunotherapy (NIR-PIT),
a treatment form that uses NIR light and an antibody photoabsorber-conjugate (APC)
to induce selective necrotic cell death in targeted cancerous cells. Our research
focuses on tracking and testing methods of evenly spreading NIR light throughout oral
tumors in order to make NIR-PIT treatment more effective. We aim to create a mathematical
model of the spread of NIR light throughout an oral tumor in order to help researchers
simulate and optimize NIR-PIT treatment for oral tumors.
Force Moment and Energy Analysis of a Three-Point Drawbar (Plough) for an Autonomous
Farm Tractor
Student: Shiva Sudireddy
Mentor/Advisor: Dr. Gemunu Happawana
Project Summary:
This project develops a force–moment-energy model for a drawbar that is to be integrated
to an autonomous farm tractor. The analytical analysis has been performed on the three-point
linkage system. This study includes black soil properties (high cohesion and adhesion,
variable moisture sensitivity) and ploughing depths ranging from 3 to 6 inches. Free-body
diagrams are prepared for the hitch, main frame, and each of the four moldboards by
resolving soil–tool interaction into horizontal, lateral, and vertical components.
A full three-point hitch analysis is then carried out to determine the two lower-link
forces and the upper-link force, along with the associated reaction moments at key
joints. The forces and moments on the linkages of the drawbar are formulated for operating
speeds up to 12 kmph, allowing dynamic load transfer to be captured during motion.
The resulting linkage forces, moments, and power demand are compared with tractor
traction and hydraulic capacity to verify safe operation and to identify critical
load cases needed for designing and adjusting the proper sensor system for the implement
that has to be integrated to an autonomous tractor.
Fresno State Rocketry NASA USLI Payload 2026
Students: Anthony Souza, John Herrera, Ryan Kane, Juan Ramos, Keagan Ott, Isaac Lay, Robert
Flath, Kooper Menefee
Mentor/Advisor: Dr. Deify Law
Project Summary:
DROP (Deployable Rover for Observation and Prospecting) is a self-deploying rover
payload that jettisons from the launch vehicle during drogue deployment and remains
tethered and protected until ground contact. After touchdown, the rover separates
from the rocket using the Tender Descender 3 (TD3) system and then drives a short
distance away from the airframe. Extendable support legs rotate the chassis into a
vertical orientation, enabling an auger-based drill to extract at least 50 mL of soil
within a 15-minute operational window. Onboard sensors subsequently measure soil pH
and electrical conductivity. All mission events are time-stamped and logged to verify
system performance, demonstrating autonomous separation, controlled descent, rover
self-stabilization, and in-situ soil analysis.
Generating Hydropower at a Non-Powered Dam
Students: Daniel S. Peckham, Alexis Publico, Angel Martinez, Marshal Spohn, Zander Clugston,
Johnathan Herren
Mentor/Advisor: Dr. Yuanyuan Xie
Project Summary:
This project will evaluate the feasibility of retrofitting hydropower generation at
the North Fork Dam near Auburn, California, an existing debris-control arch dam with
an established reservoir. The objective of this project is to determine whether a
non-hydropowered dam in California can be converted to include hydroelectric generation
while maintaining its original function and minimizing environmental and structural
impacts. The project begins with a review of the selected dam’s characteristics and
operational constraints. Multiple retrofit designs are investigated and screened based
on expected flow, constructability and anticipated financial requirements. Based on
the site analysis and the leading design alternative, a conceptual hydropower system
will be developed to evaluate turbine selection, water conveyance and routing, and
potential power output. Numerical calculations and simulations are performed to estimate
energy output at expected flow conditions. When possible, scale model testing is used
to validate assumptions and compare alternatives. The results of this project are
expected to provide a structured feasibility assessment of a potential hydroelectric
retrofit at North Fork Dam and to identify key technical and economic factors influencing
the viability of similar non-powered dam conversion projects.
Marine Energy Harvesting System
Students: Hailey Messmer, Carlos Garnica, Jonathan Gutierrez, Emilio Martinez, Isaiah Pineda-Fernandez,
Anthony Souza
Mentor/Advisor: Dr. Yuanyuan Xie
Project Summary:
This project presents the development of a compact ocean energy harvesting system
designed to generate electrical power from moving water currents. The system uses
a horizontal-axis turbine to convert the kinetic energy of flowing water into rotational
motion, which drives a generator to produce electricity. The design emphasizes a compact
and modular structure suitable for laboratory testing while representing principles
used in larger marine energy systems. Analytical modeling and simulation were used
to evaluate expected performance and guide geometry selection for stable and efficient
operation. The prototype is designed for manufacturability using additive manufacturing
methods and corrosion-resistant hardware, enabling repeated testing and future design
improvements. The system serves as a scalable platform for studying small-scale marine
renewable energy generation and improving turbine efficiency in low-flow environments.
Mobile Arm Support
Students: Ryan Rojas, Matthew Mendoza, Jose Jimenez, Zaid Srouji
Mentor/Advisor: Dr. The Nguyen
Project Summary:
The Active Mobile Arm Support (MAS) is a powered assistive device designed to help
individuals with muscular weakness improve upper-limb movement. Many current arm supports
rely on passive mechanisms, such as springs or counterweights, which can hold the
arm up but do not actively assist the user in moving. The goal of this project was
to design a table-mounted system that assists with shoulder elevation and elbow positioning
during everyday tasks.The device uses a three-degree-of-freedom linkage made from
lightweight aluminum tubing to balance strength and weight. The mechanical components
were designed using load and stress analysis, then checked using finite element analysis
to ensure safe operation. A microcontroller-based control system coordinates the motors
and power distribution to provide smooth motion.The system demonstrates reliable assisted
movement and shows how mechanical design and embedded control can be combined in a
practical rehabilitation device.
Prosthetic Hand
Students: Briyan Lucatero , Andrei Catalan, Prabal Angrish, Osvaldo Ruiz-Villegas
Mentor/Advisor: Dr. The Ngyuen
Project Summary:
This senior design project presents a redesigned below elbow prosthetic arm focused
on improving comfort, efficiency, and ease of use based on direct feedback from amputees.
Interviews with users at a prosthetic clinic in Lemoore, California, along with insights
from online amputee communities, guided the design toward increasing comfort, simplifying
attachment and reducing friction. These priorities led to a redesign of the prosthetic’s
mechanical and user-interface features to improve daily usability and reliability.
The final design is lightweight, ergonomic, and mechanically efficient, enabling smoother
motion and stronger, more dependable grasping while minimizing user fatigue. A new
attachment approach replaces traditional socket based systems to improve comfort and
reduce pressure on the residual limb. Overall, the project delivers an affordable,
user centered prosthetic solution that enhances long term wearability and quality
of life.
Robotic Helper and Guard
Students: Johan Moua, Allen Sanchez, Daniel Ornelas, Carlos Jimenez
Mentor/Advisor: Dr. The Nguyen
Project Summary:
Package theft and residential burglary impose significant financial and emotional
burdens on homeowners across the United States. An estimated 120.5 million packages
were stolen in 2023, resulting in more than $15 billion in losses, alongside more
than 839,000 reported home burglaries. This project proposes an autonomous, non-lethal
robotic security system designed to patrol home exteriors, identify unauthorized
individuals, and deter criminal activity. The system integrates four coordinated
subsystems: chassis and mobility, package handling, weapon firing, and AI-based control.
The design emphasizes affordability and practicality through an off-the-shelf tank-style
chassis, a simplified, garbage-truck-inspired robotic arm for package retrieval,
an electronically actuated gel blaster for controlled, non-lethal deterrence, and
an AI vision pipeline supported by remote GPU inference. The resulting prototype
aims to autonomously patrol, verify identity, issue escalating audio warnings, and
deploy non-lethal force as needed, providing a practical approach to reducing theft-related
losses.
Robotic Peach Harvester
Students: Natalie Maciel, Joel Velez, Carlos Vasquez, Alexander Ellis, Chathura Samarakoon
Mentor/Advisor: Dr. Ho-lung Li
Project Summary:
This project presents an AI-vision-guided robotic peach-harvesting system designed
to mitigate critical labor shortages while enhancing agricultural efficiency and sustainability,
particularly within high-density orchard environments. The proposed solution integrates
a 5-DOF robotic mechanism, featuring a specialized suction-based harvesting gripper
to eliminate fruit bruising and a custom extension arm optimized for maximum workspace
reach and mobility. To ensure precise operation, the computer-vision subsystem uses
a deep-learning architecture for real-time fruit detection and 3D localization, integrated
with the robotic arm’s motion controller via a Robot Operating System (ROS) framework.
This architecture enables autonomous path planning and adaptive grip execution based
on continuous visual feedback. The ultimate objective is to significantly reduce harvest
cycle times, minimize fruit damage rates, and provide a scalable, cost-effective framework
for transitioning to fully autonomous, intelligent orchard management in the Central
Valley and beyond.
AVI Robotics Oceanography & MATE Competition
Students: Declan Doss, Timothy Cromwell, Jaime Jaurigue, Robert Voss
Advisors/Mentors: Alaeddin Bani Milhim
Project Summary:
To address the rising demand for cost-effective subsea monitoring, AVI developed the
2026 ROV and Float platforms. These systems utilize a vectored eight-thruster array
and a custom buoyancy engine to perform high-precision maritime tasks in simulated
Arctic environments. By integrating eDNA sampling, a 7-camera optical suite, and a
robust 48V/30A power architecture, SMS has created a scalable solution for real-world
environmental protection. This project serves as a proof-of-concept for student-driven
innovation, aligning with the United Nations’ Ocean Decade goals.
Central California Engineering Design Competition Winning Design
Students: Morgan Ctibor, Alejandro Gutierrez, Madison Gonzales, Avnique Kaur Gill, Zhenhong
Li
Mentor/Advisor: Roger Moore, Eduardo Torres
Project Summary:
Introduction to Engineering students participated in the Central California Engineering
Design Challenge, which involved the design, construction, and testing of a vehicle
capable of transporting a payload of up to 50 pennies per run, representing 50 passengers.
A run was considered successful if the vehicle ascended the ramp without falling off
the ramp. Within a limited timeframe and under specified cost constraints for construction
materials, the objective was to develop the most efficient and cost-effective vehicle
possible. Multiple design concepts were evaluated, and the final model was produced
after considering key performance factors and conducting several trial-and-error tests.
The winning team’s design utilized a supercapacitor to supply power and was able to
deliver all "passengers" to the highest region of the ramp in the minimum number of
runs, earning the highest possible income score.
Sustainable and Low-Cost Housing
Students: Diego Chavez, Madison Bull, Fatima Caloca
Mentor/Advisor: Dr. The Nguyen
Project Summary:
This project involves the conceptual design and analysis of low-cost sustainable housing
in California. We are focusing on small homes between 750 and 1,000 square feet in
size. This project analyzes low-carbon materials such as bamboo, engineered wood,
sheep's wool, and carbon-sequestering concrete substitutes. This project also analyzes
alternative designs and materials for roofing, flooring, insulation, framing, and
foundations. The project deliverables will include conceptual floor plans, material
performance comparisons, and a cost analysis demonstrating how sustainable materials
can be used to develop low-cost housing.
Toy Car for Beau
Students: Gabrielle Torres, Hunter Fraser, Kaitlyn Rockholt
Mentor/Advisor: Dr. The Nguyen
Project Summary:
This project aims to design, model, and fabricate a 3D-printed toy car that meets
the needs of a six-year-old boy with KIF1A-Associated Neurological Disorder (KAND),
a neurodegenerative genetic disorder that causes visual impairment, limited mobility,
developmental delays, and other impairments. A manually pushed toy car that he will
be able to sit inside of, similar to ones made by Little Tikes or Step2, will be uniquely
designed to fit his needs, focusing on accessibility, safety, and durability. Engineering
considerations include a push handle, a lap seatbelt, a solid bottom to avoid foot
drag, and plenty of leg room. Other features include aspects like cupholders and storage
for usability. This toy car will demonstrate how engineering can be used to create
custom, accessible products that focus on user needs.
Trash Picking Robot
Students: Jake Velicescu, Ryan Caruso, and Arianna Espinosa
Mentor/Advisor: Dr. The Nguyen
Project Summary:
High school and college campuses experience persistent visible litter, contributing
to environmental degradation and increased maintenance costs. This project was initiated
to develop a fully autonomous system for pollution control on school campuses. The
system integrates a mobile chassis, mechanical collection arm, and onboard trash bin,
all controlled by an Arduino-based unit that implements motor control logic and battery-level
monitoring. The identification system consists of LiDAR sensors and a camera-based
object recognition module. System operation is supported by software implementing
obstacle avoidance algorithms and color- and shape-based trash detection. The robot
will maneuver along sidewalks and flat, open areas to identify, collect, and store
litter. The project aims to provide a cost-effective prototype to reduce visible campus
litter while demonstrating the feasibility of autonomous waste collection systems.
Trash-Picking Robot (on Water)
Students: Matthias Cockrum, Vincent Duran, Kaitlyn Bustamante
Mentor/Advisor: Dr. The Nguyen
Project Summary:
Bodies of water provide a multitude of services towards society that range from a
source of drinking water to a place for recreational activities, or to provide aesthetics
for a site. Pollution is a common issue among waterways due to human activity, usually
a result of litter or industrial activities. Pollutants such as bottles of water or
other plastics are harmful to the environment as they contaminate the water supply
and are a threat to local wildlife. This project addresses this through a robot that
functions atop water and is capable of picking up and storing trash. This is achieved
through building on a pre-built RC boat and attaching a wheel to convey trash and
store it in a bin. Sensors will indicate when the bin is full, and then be emptied.
This ensures waterways will be kept clear, and trash that is out of reach normally
can be collected.
Airborne Traffic Monitoring: Leveraging Drones for Real-Time Highway Insights
Student: Youssef Ali
Mentors/Advisors: Dr. Aly Tawfik and Roger Moore
Project Summary:
Advancements in unmanned-aerial-vehicle (UAV) technology and wireless communication
systems have created new opportunities for improving transportation monitoring and
traffic management. Traditional highway monitoring methods rely on road sensors and
fixed cameras, which often provide limited coverage and lack flexibility for rapidly
changing traffic conditions. Integrating UAV-based monitoring systems into transportation
infrastructure offers a scalable solution capable of delivering real-time aerial perspectives
across large roadway networks.
This research investigates the development of a drone-based monitoring framework designed to enhance highway traffic observation and data collection. Conducted in collaboration with the California Department of Transportation (Caltrans), the project focuses on developing UAV platforms and communication technologies capable of supporting reliable real-time traffic monitoring. The system aims to extend the capabilities of existing transportation monitoring infrastructure by enabling rapid deployment and wide-area surveillance using aerial platforms.
The research centers on the design and evaluation of a bidirectional communication
architecture utilizing Software Defined Radio (SDR) technology. This system enables
secure transmission of flight control commands, telemetry metadata, and high-resolution
video streams between UAVs and ground control stations. The study also examines the
feasibility of Beyond Visual Line of Sight (BVLOS) drone operations by analyzing regulatory
frameworks, detect-and-avoid technologies, and multilink communication strategies
necessary for long-range autonomous flight.
Additionally, the project compares the operational performance of multiple UAV platform
types—including rotorcraft and fixed-wing drones—to assess trade-offs in endurance,
payload capacity, and surveillance effectiveness. The research further explores the
integration of photovoltaic energy harvesting systems to extend UAV flight duration
and support long-term monitoring missions.
The results of this work aim to provide technical insights and prototype system recommendations
that support the adoption of UAV-based traffic monitoring technologies. By combining
advanced communication systems, aerial sensing, and regulatory analysis, this research
contributes toward the development of next-generation intelligent transportation monitoring
systems for California’s highway infrastructure.
Assessing Public Perceptions of Transportation Systems Through Multi-Modal Social Media Sentiment Analysis
Students: Anushka Patwa, Pratham Aggarwal
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
Understanding public perception of transportation systems is important for improving
transportation systems (e.g. planning, service, and commuter satisfaction). Traditional
methods such as surveys, studies, and manual feedback collection often provide limited
and delayed insights into how people experience transportation systems. In contrast,
social media platforms contain large volumes of real-time public discussions about
transportation issues such as traffic congestion, transit delays, infrastructure problems,
safety concerns, and commuter experiences. These discussions provide valuable information
that can help transportation agencies better understand public sentiments. However,
collecting and analyzing social media data for transportation research presents several
challenges, including identifying relevant posts, accessing platform data through
APIs, and selecting appropriate sentiment analysis techniques.
Prior research shows that social media can be used to study transportation issues
and traveler experiences. Researchers commonly use Natural Language Processing (NLP)
models to extract transportation-related discussions and classify them as positive,
negative, or neutral. Recent work highlights the growing usefulness of transformer-based
sentiment models for understanding context in short, informal posts.
This research uses modern NLP sentiment techniques, especially transformer-based models,
to better understand what transportation-related posts are actually saying, including
tone and context. We propose a multi-platform pipeline that collects public transportation
discussions from platforms such as Twitter (X), Reddit, Facebook, and TikTok using
API-based retrieval with keywords and date ranges. Unlike single-platform studies,
our approach supports multiple types of public data (text posts, comments, photos,
and videos/metadata where available), so we can compare what each platform captures
and reduce bias from relying on only one source. By analyzing sentiment patterns across
locations, time periods, and transportation modes, this research aims to provide a
scalable framework for monitoring public sentiment toward transportation systems and
supporting data-driven mobility planning.
EcoMaps: An AI-Powered Platform for Sustainable Transportation Decision-Making
Students: Alexander Cook, Kritika, Caleb Carillo
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
Technological innovation has largely been directed toward increasing efficiency and
profitability, with comparatively little attention given to advancing environmental
sustainability. Yet addressing the urgent challenge of human-caused climate change
requires creative applications of technology to mitigate pollution. This paper introduces
Ecomaps, a mobile application designed to empower users to make more sustainable transportation
choices.
Ecomaps integrates the Google Maps API with a machine learning–enhanced dataset covering automobile and motorcycle models from 1924 to 2024. By allowing users to enter their commute and vehicle information, the app calculates emissions, cost, and travel time across multiple transportation modes. These include personal vehicles, public options such as bus, light rail, and metro, and individual alternatives such as walking, cycling, e-bikes, and e-scooters. The interface presents comparisons in an accessible format, enabling informed decision-making that balances environmental and personal benefits.
To support user engagement, Ecomaps also features an AI-powered chatbot that clarifies uncertainties and provides tailored guidance. The app’s goal is not only to raise awareness of transportation-related emissions but also to reduce individual carbon footprints by offering actionable alternatives. Early feedback from pilot users indicates that presenting transparent trade-offs between convenience, cost, and environmental impact motivates behavior change.
By combining real-time routing, detailed emissions data, and intuitive decision support, Ecomaps demonstrates how technological tools can extend beyond efficiency and profit to serve sustainability. The project illustrates the potential of accessible, data-driven platforms to align everyday choices with global efforts to reduce pollution and combat climate change.
Evaluating the Value of Big Data for Understanding Travel Behavior to Different Land
Uses in Metropolitan Fresno
Student: Ali Abushaikha
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
The growing availability of mobility analytics and big data is transforming how cities
observe and understand travel behavior. Traditional travel surveys and manual traffic
counts often provide limited spatial and temporal coverage, particularly in small
and mid-sized metropolitan areas where detailed travel data is scarce. These limitations
make it difficult for planners to accurately understand travel demand and mobility
patterns. Large-scale mobility datasets provide an alternative by offering comprehensive,
high-resolution insights into urban travel behavior.
This study evaluates the value of big data for understanding travel behavior associated with different land uses in the City of Fresno, California. The research uses mobility data from Replica, a synthetic population-level dataset that models daily travel activity for residents and visitors. The objective is to assess how synthetic mobility data can support transportation planning and improve understanding of travel patterns in a mid-sized metropolitan city.
The analysis focuses on four destinations representing different land uses: Woodward Park (recreation), California State University, Fresno (higher education), Valley Children’s Hospital (healthcare), and the Roeding Park–Chaffee Zoo area (recreation and tourism). Using Replica’s analytical tools, custom geographic zones were defined around each site to capture trips originating or ending within those areas. Replica trip data were then used to extract key travel metrics—including trip volumes, mode shares, travel times and distances, trip purposes, and time-of-day distributions—for weekday and weekend periods.
Spatial and descriptive statistical analyses were conducted to compare travel behavior across the selected destinations. Demographic, socioeconomic, and land-use data were incorporated to examine how household characteristics and accessibility influence travel patterns.
Results show clear differences in travel behavior across land uses and notable weekday–weekend variations in trip generation and purpose. These findings demonstrate that synthetic mobility datasets can complement traditional data sources and support more informed transportation planning.
iVoxera. A Million Voices. One Intelligent Listener.
Students: Pratham Aggarwal, Kiran Nair
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
Across industries from public policy and transportation planning to healthcare research
phone surveys remain one of the most powerful tools for collecting large-scale human
insights. Organizations today face a difficult trade-off: rely on expensive human
interviewers that require weeks or months of coordination and substantial budgets,
or deploy automated IVR systems that frustrate respondents and produce low-quality,
incomplete data. iVoxera introduces a new paradigm for survey research: a fully autonomous
conversational AI phone interviewer capable of conducting natural, adaptive interviews
at massive scale. iVoxera engages respondents in fluid dialogue, dynamically interpreting
speech, clarifying responses, and guiding conversations in real time while simultaneously
capturing structured survey data. iVoxera enables thousands of simultaneous phone
interviews with 1- 2 second conversational latency. By automating the entire process,
it reduces traditional survey costs by more than 90%, while generating structured,
real-time insights.
Mobility as Shelter: Engineering a Bike-Towable Micro-Housing Solution for the Unhoused
Students: Alexa Alvarado, Aidee Gonzales, Alexander Cook
Mentors/Advisors: Dr. Aly Tawfik and Dr. Alaeddin Bani Milhim
Project Summary:
This project presents the design and development of a lightweight and modular housing
unit, initiated in partnership with the City of Kerman, in order to address the urgent
need for accessible mobile shelters for the unhoused community. Our objective is to
provide a dwelling that not only offers a practical living situation, but also offers
self-sufficiency for individuals experiencing homelessness while ensuring mobility,
safety, accessibility, and affordability. This project demonstrates how community-driven
engineering, grounded with compassion and practicality, can deliver humane, affordable,
and sustainable solutions to homelessness at both the localized and systemic levels.
Path: Practical Gen-AI Tech for Transportation Hub
Students: Aidee Gonzales, Pratham Aggarwal, Anushka Patwa, Cesar Aranibar, Kritika, Alexander
Cook, Youssef Ali, Andres Ramos
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
This study evaluates participant experiences and learning outcomes from a full-day
Generative AI workshop designed for transportation and public-sector professionals.
The workshop consisted of four interactive sessions introducing foundational concepts,
hands-on demonstrations of AI tools, technical implementation strategies, chatbot
development, and responsible AI practices. Pre- and post-workshop surveys were used
to assess participants’ familiarity, confidence, and perceptions of AI before and
after the training.
Paving Pathways to the Future: Fresno State Transportation Institute’s K-12 Summer
Camps for Transportation, Sustainability, and STEM
Student: Cesar Aranibar
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
The Fresno State Transportation Institute provides early exposure to transportation,
sustainability, and STEM through summer camps for elementary, middle, and high school
students. The program helps students understand transportation’s role in safety, mobility,
infrastructure, creativity, and emerging technology while introducing future educational
and career opportunities.
Camps offer age-appropriate, hands-on learning connected to real-world applications. Transportation Art Camp introduces elementary students to transportation through creative design, vehicle modeling, and public art. Transportation Academy and Bike Camp engage middle school students in transportation systems, sustainability, bicycle safety, maintenance, and riding, with activities such as 3D printing, coding robots, and magnetic levitation. AI in Transportation Camp and Drone Camp serve high school students through applied learning in object detection, smart mobility, aerial mapping, drone safety, autonomy, and infrastructure problem-solving.
Assessments show positive growth in transportation knowledge, career interest, teamwork, and applied problem-solving.
Riding on Data: AI-Based Modeling of Motorcycle Emissions
Students: Kritika, Andres Ramos, Alexander Cook
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
Transportation is a major contributor to climate change, primarily through the emission
of greenhouse gases, pollutants that trap heat in the atmosphere and drive global
warming. The most significant greenhouse gas from vehicles is carbon dioxide (CO2),
alongside methane (CH4) and nitrous oxide (N2O). Motorcycles represent one of the
transportation segments, yet accurate emission prediction models for these vehicles
remain limited in existing transportation planning tools. This research presents an
AI-modeling approach for predicting motorcycle greenhouse gas emissions from readily
available vehicle specifications.
The model utilizes EPA motorcycle certification data spanning 2006 to 2025, encompassing comprehensive technical specifications and emission measurements for thousands of motorcycle models. The dataset includes engine displacement (cubic centimeters), power output (horsepower), cylinder count, and emission control technologies, specifically the presence of catalytic converters and Exhaust Gas Recirculation (EGR) systems. The target variables are different greenhouse gases, representing actual laboratory-measured emissions in grams per kilometer.
Multiple AI, specifically machine learning, algorithms like SVR, NN, Rule Model, linear regression random forest and KNN were evaluated during model development, with careful comparison of their predictive capabilities. The selected approach demonstrated strong performance in handling non-linear relationships between vehicle specifications and emissions, managing multicollinearity among features, and providing interpretable insights into which factors most influence emissions.
Users of this model receive accurate, vehicle specific emission estimates rather than generic averages when planning motorcycle trips. Transportation planners can assess the environmental impact of different motorcycle fleets and identify high emission vehicles for targeted interventions. By providing transparent, accurate predictions based on readily available vehicle specifications, this model transforms abstract environmental concerns into concrete, actionable information that supports evidence-based decision-making for individuals, organizations, and governments working toward transportation sustainability goals.
TransitHub-AI-Powered Geospatial Platform for Transit Accessibility and Equity Analysis
Students: Pratham Aggarwal, Evan Solis
Mentor/Advisor: Dr. Aly Tawfik
Project Summary:
Advancements in AI and geospatial analytics offer new opportunities to address transportation
equity. This research introduces TransitHub, a novel AI-powered platform designed
to visualize and assess transit accessibility across the United States. Unlike traditional
tools relying on static schedules, TransitHub leverages Google APIs to harvest real-time
performance metrics for millions of addresses, fusing them with census-level socioeconomic
data. The platform features a “drill-down” interface utilizing K-Means clustering
to automatically segment regions into performance quadrants, effectively identifying
“transit deserts” and “equity gaps.” Furthermore, it integrates a Retrieval-Augmented
Generation (RAG) architecture and a Planner-based AI chatbot, enabling users to query
complex database schemas using natural language to generate instant, data-backed insights.
By bridging the gap between abstract statistics and actionable policy, TransitHub
provides a scalable decision-support system for optimizing infrastructure planning
nationwide.
Uncovering California’s Most Dangerous Roadways: An Analysis of Fatal Crash Rates
in Every County Across the State
Student: Alexis Perez
Mentors/Advisors: Dr. Aly Tawfik, William Bommer
Project Summary:
Transportation is essential to economic and social activity, but roadway safety remains
a major concern due to persistent traffic fatalities. This project identifies and
analyzes California’s most dangerous road segments based on fatal crash rates to support
targeted safety improvements. Using crash data from the Fatality Analysis Reporting
System (FARS) and traffic volume data from INRIX, the study evaluates risk through
measures such as crashes per mile per year and per million vehicle miles traveled
(MVMT). To strengthen the analysis, spatial methods including 200×200 ft grid cells
and a decaying time metric were applied to identify consistent high-risk areas and
patterns over time. The findings provide actionable insights for agencies and policymakers
to prioritize interventions, improve roadway design, and implement data-driven strategies
to reduce fatal crashes across California.
An Agent-Based Model for the Analysis of Groundwater Usage and Replenishment
Students: Jaden Luna, Jose Eduardo Rodriguez
Mentor/Advisor: Dr. Jorge Pesantez
Project Summary:
Groundwater supplies roughly 40% of the water used by Californians, rising to 60%
during drought years. This reliance on groundwater can lead to overdrafting, resulting
in dry wells, subsidence, pollution, and more. These effects are limited by groundwater
replenishment, where water is pumped back into aquifers. This study focuses on modeling
groundwater consumption and replenishment with an agent-based model (ABM) approach.
An ABM simulation enables a comprehensive evaluation of groundwater demand and supply.
The model planned by this study would be able to identify trends in groundwater users
represented by agents, potentially showing early signs of overdrafting, and replenishment
efforts may be needed. Results are expected to inform users about where and when high
consumers might occur and identify areas where groundwater issues could arise due
to overdrafting.
Deep Learning-Based Geometric Characterization and 3D Localization of Oranges and
Stems in Field Environments
Students: Andrei Catalan, Kinnoree Pasha
Mentor/Advisor: Dr. Alaeddin Bani Milhim
Project Summary:
Complex agricultural environments, including fruit overlap, leaf occlusion, and variable
lighting, limit the effectiveness of traditional image processing methods. This research
implements an RGB-D camera and YOLOv8 instance segmentation to improve citrus fruit
and stem recognition in field conditions. Segmentation masks are post-processed using
contour-based centroid extraction to obtain stable pixel targets, which are converted
to metric distances using depth measurements and calibrated camera intrinsics for
3D localization. A dataset of 1,159 images collected at Fresno State farms was augmented
to 3,391 images using brightness, saturation, and exposure variations, and manually
labeled into orange and stem classes for supervised training. On the held-out test
set, the model achieved a mean average precision at 50% overlap of 88.0%, with 99.0%
for oranges and 76.0% for stems, demonstrating robust fruit segmentation and reliable
stem discrimination for autonomous harvesting applications.
DeltaX Robot Autonomous Weed Detection and Elimination for SARDOG
Students: Daniel Remington, Armanjit Gill, Brandon Wilbur
Mentor/Advisor: Dr. Hovannes Kulhandjian
Project Summary:
The DeltaX Weed Removal robot was developed to support Fresno’s agrarian community.
Its primary goal is to remove weed pulling, a repetitive and labor-intensive task,
so farmers can devote their time to other important duties on the field. The robot’s
main function is to detect and remove weeds using an AI-based vision system and a
flexible robotic DeltaX arm. The AI model will be trained on a large sample of pictures
of weeds native to Fresno, California. DeltaX is controlled via a Raspberry Pi using
a ROS2 framework. ROS will be programmed to coordinate navigation, image processing,
and motor movement. The project will involve a DeltaX arm, a CPU coordinator (Raspberry
Pi/Thor Jetson), the YOLOv9 AI detection system, and a MiDaS depth-detection model
paired with ROS2. Our project aims to demonstrate the positive impact that robotics
can have on agricultural productivity.
Development of an Electrodialysis-Based Water Treatment System for Selective Ion Removal
and Recovery of Phototoxic Contaminants
Students: Hovhannes Torikyan, Gurkeerat Singh, Miguel Sanchez, Dominic de Castro
Mentors/Advisors: Dr. The Nguyen, Dr. Sankha Banerjee
Project Summary:
This project evaluates, automates, and optimizes an electrodialysis reversal (EDR)
system for precise ion separation from saline and brine streams. EDR is a membrane-based
electrochemical process that selectively transports ions across alternating cation-
and anion-exchange membranes under an applied electric potential, producing dilute
and concentrate streams. The work focuses on improving reliability, controllability,
and experimental repeatability through system-level enhancements. A key objective
is developing an automated control architecture using a programmable logic controller
(PLC) and operator interface to regulate pumps, valves, and system monitoring. This
automation aims to improve safety, scalability, and consistency while reducing manual
intervention. Concurrently, computational fluid dynamics (CFD) simulations using ANSYS
tools and OpenFOAM examine transport behavior within the hydraulic loop and membrane
channels, identifying inefficiencies such as flow nonuniformity or pressure losses.
Integrating experimental operation with simulation-guided design adjustments supports
development of a more robust, efficient, and scalable EDR platform.
Precision Citrus Harvesting Robotic System for Controlled Stem Length
Students: Brendan Chinnock, Andrei Catalan, Alexander Ellis, Daniel Yang, Nicole Oliva, Juan
Espinoza, Caleb Banhe
Mentor/Advisor: Dr. Alaeddin Bani Milhim
Project Summary:
This project focuses on the development of a precision citrus harvesting robotic system
capable of cutting fruit with a controlled residual stem length. Controlled stem length
is critical to maintain the fruit quality and market value during post-harvest handling
and distribution. The system addresses increasing agricultural labor shortage and
improves harvesting efficiency. The system consists of a mobile platform, a 6-DOF
Fairino robotic arm, an RGB-D camera, and a custom-designed end-effector. This end
effector incorporates three TPU-filament gripper fingers for compliant fruit grasping
and an integrated shearing mechanism for controlled stem cutting. Using a machine
learning-based fruit recognition system, the robotic manipulator positions the end-effector
in contact with the target fruit. A novel motion strategy incorporating lateral and
vertical vibrations is applied to align the stems within the shears blade for precise
cutting. Once the stem is cut, the fruit is automatically harvested and deposited
into a collection bin. Field experiments validated the system’s performance, achieving
a 95% harvesting success rate with an average cycle time of 40 seconds per fruit.
The residual stem length consistently met the 2-mm target, satisfying post-harvest
quality requirements.
Remote Sensing-Based Leak Detection in Agriculture
Student: Aaron Millwee
Mentor/Advisor: Dr. Alaeddin Bani Milhim
Project Summary:
Water loss from undetected leaks in agricultural irrigation systems poses significant
economic and environmental challenges. Irrigation pipes and hoses are often shallow
or surface-level, vulnerable to damage from equipment, and difficult to monitor across
systems that may span hundreds of acres. This research reviewed non-invasive water
leak detection and localization methods, with emphasis on acoustic techniques, ground-penetrating
radar (GPR), infrared (IR) thermography, and microwave remote sensing. Acoustic approaches,
including leak noise correlation and hydrophone-based detection, effectively identify
leak-induced vibrations in plastic pipes but require close proximity to the pipeline,
limiting their feasibility for large-scale agricultural deployment. GPR detects leaks
by exploiting dielectric contrasts associated with soil moisture anomalies; however,
its performance is highly dependent on soil composition, moisture variability, calibration,
and ground proximity, making it impractical for aerial platforms. IR thermography
enables rapid, wide-area monitoring by identifying surface thermal anomalies, but
its reliability is affected by environmental factors such as solar radiation, soil
moisture conditions, and vegetation cover, leading to false positives when used independently.
Passive microwave remote sensing provides broad soil moisture measurements but lacks
the spatial resolution required for precise leak localization. Overall, the literature
indicates that no single method can provide reliable and cost-effective detection
across diverse agricultural conditions. Future work will investigate a hybrid approach
combining drone-mounted, low-cost IR imaging for wide-area detection with localized
soil moisture verification through targeted drilling and moisture sensing performed
by the same drone at suspected leak locations.
CTEC High School
Bio-Medical Engineering: Improving Medical Ports - Impossible or Not?
Student: Zaire Woodruff
Mentor/Advisor: Bryan Sheldon
Project Summary:
For Fresno State's 2026 LCOE project day, I have researched and collected information
on the medical device aspect of biomedical engineering by learning and educating others
on medical ports. Going into this research, I knew very little about medical ports,
so I wanted to learn more to learn the specifics on small details to conclude whether
or not things can be changed or added to the design. Most existing research is what
they are and why/who they are intended for, but for my project, I tried to find the
history and the “whys” of their particular design choice. I chose medical ports because
I wanted to give insight and information about myself and people living with medical
devices like mine, and I chose to dive into details because I want to find a way,
if there is a way, to improve them.
Delta-style Low-cost 3D Printer Creation
Student: Nathan Blodgett
Mentor/Advisor: Bryan Sheldon
Project Summary:
The goal of this project is to construct an affordable, scalable, delta-style FDM
3D printer with modern features. 3D printing is an industry that has been undergoing
various innovations over the years, including the use of delta-style kinematics, probing,
better firmware, superior hotend and extruder performance, and other advancements
that have become common and cheaper. In all DIY 3D printers, it should be noted that
3D printed components quite often have other components, such as heat-set inserts
and plenty of screws. This project is made with easily sourced parts and mostly 3D
printed parts (from a printer with a 220mm buildplate), and takes advantage of the
pioneering work done in previous projects. This project is meant to be easily accessible
to an audience who wants to have the freedom to easily modify their own printer.
Predicting Object Trajectory: Mathematical Modeling and Engineering Applications
Student: Shealyn Dostalik
Mentor/Advisor: Bryan Sheldon
Project Summary:
Trajectory prediction is essential in engineering, robotics, and physics because it
gives accurate prediction of where a moving object will travel and land. This project
explores the mathematical foundations of trajectory prediction, focusing on how variables
such as velocity, launch angle, gravity, rotational axes, and external forces shape
an object’s path. Using calculus-based modeling, CAD tools, and simulation software,
the research demonstrates how motion across multiple axes can be predicted with high
accuracy while accounting for potential human error.
The project will conclude with a research display and live demonstration showcasing real-world engineering applications, including robotics used in competitive design. By linking historical developments with modern industry practices, this study connects multidimensional mathematics to practical solutions. Ultimately, the project highlights the complexity behind mechanical motion and emphasizes the importance of precise calculations while developing technical skills in analysis, problem-solving, and engineering documentation for future study in civil engineering
Properties of Aerodynamics and Air Flow
Student: Michael Rodriguez
Mentor/Advisor: Bryan Sheldon
Project Summary:
This project investigates the basis of aerodynamic forces and the flow of air within
a controlled environment. This will be modeled with a wind tunnel built to be used
to study and visualize the flow streams of air in a controlled environment and how
foreign objects change the flow. The results from this study will contribute to a
better understanding of the relationship between the shape of objects and aerodynamic
performance, offering insights applicable to fields such as aerospace engineering,
automotive design, and fluid dynamics research. The design and creation of a wind
tunnel also gave research and an understanding of how air behaves and some functions
of it that need to be considered within aerospace engineering.
Prosthetic Arm
Student: Denali Singleton
Mentor/Advisor: Bryan Sheldon
Project Summary:
When examining the source, I believe the plan of action would have to choose between
three options for prosthetic arms. The options are Passive, Body-powered, Myoelectric,
and Hybrid, but the first option that is out is Passive. This leaves body-powered
or myoelectric prosthetic arms, but I would opt for the myoelectric option because
it has more potential power than the other option. It had more ways of building than
the other one. When I present my project as an end product, I will focus on the way
I control the arm and the arm itself, but that doesn't mean I won't also review how
I built the arm. It would use a 3d printer for most of the build. A low-cost prosthetic
arm that can handle simple daily tasks for individuals without an arm, allowing them
to perform activities they couldn't do before.
Edison High School
BiteBot: Just Put it in the Pot
Students: Sullivan Peebles, Liam Perez, Christopher Shabaglian, Emerson Fabian-Hicks
Mentor/Advisor: Xiong Cha
Project Summary:
BiteBot is an automated food preparation system designed to simplify cooking through
intelligent robotics. By placing ingredients on a designated tray, users enable BiteBot
to identify, handle, and assemble food items without manual intervention. The system
uses a robotic arm to precisely pick, move, and manipulate ingredients according to
predefined preparation steps, ensuring consistency and efficiency. BiteBot aims to
reduce human effort, minimize preparation time, and improve hygiene in food handling.
Its design demonstrates the practical application of robotics and automation in everyday
tasks, highlighting how smart systems can enhance convenience and reliability in the
kitchen.
Dustless Almond Harvesting
Students: Adan Gutierrez Lopez, Adrian Ortiz, Albert Tejeda, Jorge Lara Torres, Stephen Barkema
Mentor/Advisor: Xiong Cha
Project Summary:
Almond harvesting in California’s Central Valley is a significant source of particulate
matter (PM2.5 and PM10), contributing to some of the worst air quality in the United
States and increasing the risk of respiratory and cardiovascular disease in surrounding
communities. Although low-dust harvesting equipment exists, adoption remains limited
due to high upfront costs, equipment replacement requirements, and time constraints
during harvest season. This project proposes the development of a low-cost, high-efficiency
dust-reduction system designed to integrate with existing almond harvesting equipment.
By prioritizing affordability, ease of installation, and minimal operational disruption,
the proposed solution addresses the primary barriers preventing grower adoption. The
system aims to significantly reduce harvest-generated dust while remaining economically
viable for small and mid-scale growers. Successful implementation would improve regional
air quality, protect public health, and support sustainable agricultural practices
without imposing additional financial burdens on farmers.
Ecosoak
Students: Angel Perez, Jovanni Manzo, Jose Carlos, Real Vang, Selena Wall
Mentor/Advisor: Xiong Cha
Project Summary:
Working-class individuals often lack the time required to maintain a healthy residential
garden due to demanding work schedules and daily responsibilities. Traditional gardening
requires consistent attention, including watering, monitoring soil conditions, and
managing plant health, which can be difficult to sustain. This project proposes a
low-maintenance, time-efficient gardening system that reduces the need for constant
human involvement. The solution incorporates automated irrigation and environmental
sensors to monitor factors such as soil moisture and temperature, enabling real-time
adjustments to plant care. Designed with affordability and ease of use in mind, the
system aims to be accessible to a wide range of users. The expected outcome is improved
plant health and productivity with minimal time investment, promoting sustainable
living and making home gardening more practical for busy individuals.
HestiaLck
Students: Jesus Diaz, Jason Jordan, Daniel Green, Joesph Hart
Mentor/Advisor: Xiong Cha
Project Summary:
In May 2025, a widely circulated news report detailed how a 2-year-old in Philadelphia
accidentally shot himself after discovering an unsecured handgun in a bedroom. The
incident, marked by its ordinary circumstances and complete preventability, prompted
our group to ask our instructor whether unintentional child firearm injury could serve
as a meaningful focus for our project. From there, we conducted independent research
to determine whether this event was an isolated tragedy or part of a broader national
pattern.
The data we examined made the scope of the issue immediately clear. National surveys indicate that over half of gun-owning households keep at least one firearm unlocked, and more than one-fifth store a loaded gun without any safety device. Multiple studies further show that children frequently know where firearms are stored and can access them within minutes, even when parents assume otherwise. These conditions directly mirror those that led to the Philadelphia shooting.
Our understanding solidified upon reviewing a CDC analysis of the National Violent Death Reporting System (NVDRS). The report documented more than a thousand unintentional firearm deaths among children over nearly two decades, with the majority occurring in the home and involving firearms that were both loaded and unsecured. The NVDRS data provided definitive evidence that these incidents are not random or rare; they follow consistent, preventable patterns.
Taken together, the recent news event, supporting national data, and federal reporting systems confirmed that unintentional firearm injury among children represents a significant and ongoing public safety problem. This evidence formed the basis for selecting it as the focus of our project.
Navissist
Students: Enrique Garcia, Ethan Saporta, Anthony Espinoza, Lane Loucks
Mentor/Advisor: Xiong Cha
Project Summary:
Visually impaired people have trouble navigating through packed urban areas that are
riddled with hazards and weren't designed for them; without assistance these people
suffer from a lack of independence and increased social isolation.There is a large
population of blind (those with lack of sight) people in the US with the count being
around 1 million and even more, at around 3 million, being considered visually impaired--
of these visually impaired people, most are of lower income levels and people who
lack private health insurance as well as non--white populations. While existing remedies
to the issues exist like seeing eye dogs, they require training, constant care, and,
most importantly, a lot of money at around $50k. Most people cannot afford human assistance
and rely on their friends and family to help them, for those who live alone, this
is not an option.The people most at risk and affected by visual impairment do not
have the resources to sustain themselves, let alone a seeing eye dog or many forms
of assistance. There is a need for an affordable, low maintenance, reliable assistant
to help tackle the issue of navigation through urban environments. Our device is a
clip on sonar sensor that is an affordable option to solve the problem above.
OpenTheWay
Students: Emily Barrios-Torres,Angelo Padre,Victoria Balderas,Cassandra Navarro,Yamileth Carranza
Mentor/Advisor: Xiong Cha
Project Summary:
This project, OpenTheWay, focuses on designing front doors that are more accessible
and convenient for people who have upper impairments and arm disabilities, causing
them to encounter many challenges when opening doors. We would like to create more
independence for those who want it, but are limited due to their disabilities. Additionally,
we wanted to create a more inexpensive product that allows people who can not afford
a regular, residential door opener to gain access to these automatic operated devices.
With this product, it would allow the user to have more ease when opening doors with
no or little assistance.
Scaffold Safety
Students: Jerry Naviz, Manuel Ortigoza, Robert Cortez, Mario Gonzalez
Mentor/Advisor: Xiong Cha
Project Summary:
Our project focuses on the halting of scaffolding related accidents in a construction
site. With our research proving that most accidents happen due to poor assembly, human
error, and the inadequate use of safety protocols. We plan to design and integrate
a scaffolding that we can use for the future. Our scaffold is to be OSHA approved,
as well as improving those OSHA rules, and follows all guidelines that need to be
set in a construction site. This solution is designed to be compatible with existing
scaffolding infrastructure, and compliments with current safety standards, ultimately
creating a safer and more reliable working environment for construction professionals.
Viewpoint-Consistent Image Recovery for Monitoring Historic Structures Using Photogrammetry
and 3D Gaussian Splatting
Student: Sullivan Peebles
Mentor/Advisor: Xiangxiong Kong
Project Summary:
Historic structures and sites are essential cultural assets that embody community
identity and irreplaceable craftsmanship. Maintaining their integrity over time requires
repeatable inspection methods that can detect subtle changes. In this work, we propose
a novel image-recovery framework to enable consistent viewpoint comparison for vision-based
monitoring of historic structures across multiple inspection periods. Models generated
from photogrammetry and 3D Gaussian Splat are constructed from a new image campaign,
and an old archival photograph is aligned to 3D models to recover its camera pose.
The recovered camera parameters are then imported into a computer graphics environment
(e.g., Blender) to render synthetic images from the same viewpoint of 3D models but
reflecting the structure’s updated condition. This produces temporally comparable
image pairs without requiring repeat photography from identical camera positions.
Where's My Water
Students: Josue Tepec, Robert Ventura, Juan Santiago, Spencer Setha, Enrique Garcia
Mentor/Advisor: Xiong Cha
Project Summary:
Our team is creating a cost effective app integrated with sensors ditributed throughout
irrigation lines. This app will read presure data thorugh the irrigation system and
detect any loss of pressure and mark as a leak allowing for the farmer to go and fix
rather than be fined for excess water ussage.
Sanger High School
Functionalized Bio-composite Sorbent: Novel Core-Shell beads with Encapsulated Functionalized
Biochar Engineered for Selective Chemical Removal of Waterborne Contaminants, Affording
Magnetic Retrieval for Optional Circular Water Management
Student: Julianne Luna
Mentor/Advisor: Davin Aalto
Project Summary:
Widespread contamination of surface and groundwater systems has intensified the need
for sustainable water remediation technologies. Previous experimentation suggested
success in removal of generalized pollutants; this project sought to develop a functionalized,
multi-layered, magnetically retrievable bio-sorbent for high specificity capture in
contaminated aquatic environments, facilitating a circular water management system.
Core-Shell beads were engineered for selective absorption based on various chemical
mechanisms, promoting selective affinity. Three pollutant types were selectively targeted:
heavy metal micronutrients (Recipe A), toxic heavy metals (Recipe B), inorganic plant
macronutrients (Recipe C). Core-Shell beads in solutions were assessed pre and post
84 hour exposure, measuring concentration differences, ensuring selective removal.
Recipe A reduced pollution levels in Copper Sulfate by 55%, Recipe B reduced Zinc
concentrations by 40%, Recipe C reduced Phosphate levels by 57%. In the Well Water
Test, Recipe A removed 100% copper ions, Recipe C removed 67% nitrate ions and 62%
overall ions. The findings and analysis suggest the success and further potential
of an engineered magnetically retrievable, functionalized, core-shell structure selective
for targeted contaminants as a more efficient circular water treatment approach.