Faculty: Lubos Brieda, AERO

Email: lbrieda@calpoly.edu 

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Densities, temperatures, potentials, and energies in plasma thruster plumes are characterized using Langmuir, Faraday, and Retarding Potential Analyzer probes. The objective of this summer research is to design, build, and test a small sensor suite containing these three probes that will be used in Cal Poly’s new 6×14 foot STELLA vacuum chamber. The participant will also construct an electrical harness, readout circuitry, and a simple LabView screen to aid in data collection. Activities will be summarized in a conference paper. The sensor pack developed as part of this SURP will be utilized by future graduate students and external clients performing plasma propulsion testing in this facility.

Faculty: Nandeesh Hiremath, AERO

Email: nhiremat@calpoly.edu 

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

The project aims to establish a table-top heat transfer experiment. Advancing the aerospace curriculum in these areas will lay the foundation for undergraduate students pursuing experiential learning. A specific laboratory experiment involving conduction, forced and free convection, will be scoped, designed, and fabricated during the SURP performance period. The outcomes of this work will standardize the introductory courses on heat transfer specific to aeronautics and astronautics concentrations, which are mainly lacking outside of the Cal Poly Aerospace curriculum in an undergraduate setting. The work will also address the pedagogical methods that adhere to Bloom’s Taxonomy, meeting learning outcomes, grading, and assessment. The student researcher will gain hands-on experience with conducting the specified activities, developing an operational and safety manual, and potentially co-authoring an education research-based paper.

Faculty: Nandeesh Hiremath, AERO

Email: nhiremat@calpoly.edu 

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Autonomous flight systems require precise tuning of control inputs to account for changing flight conditions. The project aims to achieve an autopiloted flight system for a high-aspect-ratio wing designed for high-altitude, low-density flight conditions. The project scope is to integrate and assemble the flight controller and autopilot system into a flight demonstrator within the permissible weight category for consumer UAVs. This would include pre-programmed flight path planning, implementation of flight mechanics, and trim based on the onboard inertial measurement units. The desired control inputs will be developed using (a) a tabletop setup of a scaled wing and control surface with integrated autopilot modules, (b) followed by integration to the flight platform. This entails limited wind tunnel and flight testing.

Faculty: Stephen Kwok Choon, AERO

Email: skwokcho@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This Summer Undergraduate Research Project focuses on the proposed design, development, manufacturing, and preliminary testing of a Reaction Wheel for use on a Floating Spacecraft Simulator test bed. This project aims to design, build, and test a reaction wheel module that shall be installed on a Floating Spacecraft Simulator vehicle to be used for a teaching laboratory experiment and research applications in the Space Robotics Laboratory and classroom (AERO-471 Introduction to Space Robotics – Lab).

Faculty: Madhusudan Vijayakumar, AERO

Email: msudan@calpoly.edu

Accepted Projects Modes: Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

With the cost of launch becoming more affordable, Low Earth orbit (LEO) is becoming increasingly congested, elevating the importance of accurate orbit prediction for space traffic management. However, current operational orbit prediction methods depend on uncertain or unavailable spacecraft metadata, time-varying environmental effects, and limited knowledge of maneuver intent. Further, publicly available orbit catalogs provide broad coverage but are noisy and irregularly sampled. This project aims to develop a scalable, maneuver-aware orbit prediction pipeline that learns from historical TLE to predict nominal orbital evolution over 48-hour window and flags potential maneuvers providing an associated reliability measure. Using publicly available TLE for congested LEO regimes, physics-based simulation for baseline comparison and controlled maneuver injection, we will create a curated dataset. By developing a time series neural network, we will benchmark prediction performance and evaluate maneuver detection on both synthetically labeled cases and known maneuvering objects such as the ISS. The outcomes of this project will be integrated into PolySpace, Cal Poly’s mission design and operations platform enabling reproducible research in next-generation space traffic management.

Faculty: Christopher Heylman, BMED

Email: cheylman@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Our lab is developing a microfluidic “tumor-on-a-chip” device that will allow for the growth and maintenance of 3D vascularized human colorectal cancer tumors. These tumor tissues will be used for screening the effects of novel drugs on human colorectal cancer before resorting to costly pre-clinic animal models and human clinical trials. These devices are created by injecting a mixture of human fibroblasts, endothelial cells, colorectal cancer cells and extracellular matrix proteins into a central incubation chamber in a microfluidic device. Cell culture medium is then perfused through the tissue using the fluidic channels of the device. Given the appropriate ratio of cell types, nutrients in the medium, and flow rates, a 3D tumor with an integrated network of blood vessels can be grown. This summer research project aims to identify and develop new assays for assessing the metabolic state of these tumors on a chip. The skill sets that students will develop in this project include human cell culture, molecular assays, fluorescent microscopy, and image analysis using ImageJ software. Successful completion of the aims of this project will open the door for further research and use of these devices to screen drugs for efficacy in treating colorectal cancer. Establishing assays for quantitative analysis of these tumor tissues will increase the functionality of these devices as a tool for the pharmaceutical industry.

Faculty: Shreeshan Jena, BMED

Email:shjena@calpoly.edu
Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

The present proposal (SURP 2026) targets the utilization of the biomimetic model of a foot developed from previous research (SURP 2025) and to implement the forces experienced by human foot under various activities of daily living. The incidences of foot pain, acute or chronic, can be studied in detail, with the use of proper biomimetic models of the human foot [1, 2]. The loads experienced by the tendons at the different sub-phases of the stance phase of human gait will be computed using the computed muscle control tool (CMC) available through OpenSim, and the resultant stresses on the bone, cartilage, and soft-tissue geometries will be evaluated using a finite element analysis platform. An inertial motion unit (IMU) based motion capture system will also be developed to record new lower limb gait kinematic data and implement the recorded postures during the corresponding sub-phases, on the biomimetic model. The stresses and deformation distributions on the model will be studied by comparing data recorded from instrumented insoles, and the Kite’s and the Costa-Bartani angles [1]. The successful validation of the CalPoly foot biomimetic model would be instrumental for future studies under different gait or pathological conditions.

Faculty: Thomas Katona, BMED

Email: tkatona@calpoly.edu

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This summer research experience for undergraduates offers an opportunity to investigate the effects of light on human physiology and health through both literature-based and experimental approaches. The student will conduct an in-depth review of current scientific research on how light exposure— including intensity, duration, timing, and spectral composition— influences human health outcomes such as circadian rhythms, sleep cycles, alertness, and overall well-being. In parallel, the student will systematically tear down and analyze commercially available lighting products that claim health benefits, as well as products suspected of having detrimental health effects, to characterize their optical properties and performance. Experimental measurements of light intensity and spectrum will be compared against established findings in the scientific literature to evaluate the validity of marketing claims and potential health implications. This project provides hands-on experience in scientific research, critical analysis of health-related claims, optical measurement techniques, and the integration of experimental data with peer-reviewed literature.

Faculty: Luke Perreault, BMED

Email: lperreau@calpoly.edu

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Tissue engineering has revolutionized fields such as biomedical research and food systems (cellular agriculture/lab-grown meat), yet bioengineering topics are rarely incorporated into secondary STEM curricula. Further, engineering topics taught in schools are often isolated from ethical and societal contexts, despite evidence that student interest in STEM increase when learning is grounded in real-world challenges emphasizing relevance, sustainability, and social impact. Integrating bioengineering into secondary education is particularly challenging due to limited resources and insufficient teacher preparation. This project addresses this gap by developing engineering design modules leveraging low-cost plant-based biomaterials under development in my lab. The SURP student will work at the intersection of biomaterials research and engineering pedagogy, training in plant decellularization and scaffold characterization techniques such as mechanical testing and microscopy. These laboratory experiences will be translated into classroom-ready design challenges that emphasize creativity, agency, and design for diverse communities. The student will integrate human-centered engineering design principles into lesson plans and protocols and pilot activities through summer outreach programming at Cal Poly. Through this work, they will gain experience in biomaterials research and inclusive engineering education while contributing to scalable pathways into bioengineering for historically underserved youth. This initiative aligns with CENG’s commitment to inclusive excellence by expanding Learn by Doing opportunities that integrate equity, social impact, and engineering practice.

Faculty: Luke Perreault, BMED

Email: lperreau@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

The demand for transplantable organs and tissues vastly exceeds availability, and an average of 22 people die each day while waiting for a donor. Tissue engineering offers a promising method to increase the availability of viable tissues for patients by growing cells within biomaterial scaffolds designed to mimic native tissue structure. However, current scaffold fabrication methods are costly and struggle to replicate the vascular networks required for cellular oxygenation. Nature offers an elegant solution: through decellularization with common detergents, our lab can repurpose plants into cost-effective, sustainable scaffolds with intricate structures and vasculature networks. Despite these advantages, plant-derived cellulose lacks mammalian cell-binding motifs, requiring surface modification to support cell adhesion and survival. In collaboration with Dr. Morgan Hawker of CSU Fresno, a specialist in biomaterial plasma modification, we will evaluate nitrogen plasma surface activation on decellularized spinach leaf scaffolds as a cost-effective strategy to improve scaffold biocompatibility. The student researcher will prepare plant scaffolds, quantify plasma-induced changes in surface wettability and protein adsorption, and evaluate cell adhesion and viability using fluorescence microscopy. By developing a scalable method to enhance cell integration on plant-based scaffolds, this project advances sustainable biomaterials toward therapeutic relevance within a cross-CSU, Learn-by-Doing, tissue engineering research experience that bridges kingdoms of life to improve human health.

Faculty: Soph Ziemian, BMED

Email: sziemian@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Female athletes have a 2–8 times higher risk of anterior cruciate ligament (ACL) injury compared to male athletes. Sex differences in ACL injury incidence are largely attributed to differences in movement mechanics, anatomical structure, and hormonal influences. This project will be a pilot data collection and validation study examining the kinematics and kinetics of dynamic movements specifically linked to ACL injury risk. The data will be collected in the Mobile Biomechanics Lab using both traditional marker-based motion capture (Cortex) and a mobile markerless method (OpenCap). A SURP student will collect synchronized trials across multiple dynamic movements (i.e. jumping, landing, squatting, and change-of-direction tasks), perform kinematic and inverse dynamic analysis of data, and quantify agreement and reliability across methods. The main scientific objective is to assess if OpenCap-derived measures will be sufficiently accurate and sensitive to detect sex differences in college-aged participants in mechanics across these movements. Results will directly inform the feasibility, protocols, and analysis strategies for a planned larger-scale study to assess sex differences in dynamic movements associated with ACL injury risk in youth athletes. This project will also work towards longer-term goals of improving scalable mobile biomechanics assessment methods for research and building community outreach efforts that connect youth athletes with biomechanics to promote engagement and sustained involvement in STEM.

Faculty: Jiun Yao Cheng, CEEN

Email: jcheng95@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Artificial Intelligence (AI) is poised to transform the construction industry by reshaping work processes and environments. With these disruptive changes, a significant gap exists regarding how higher education institutions could better prepare future workforce in response to the potential opportunities and concerns that AI could bring to the industry. This project serves as the foundational phase of a larger effort to foster an “AI-Ready” workforce, aiming to investigate domain-specific AI literacy and attitudes among key stakeholders in the construction engineering community. By identifying discrepancies in AI usage and literacy, this research will provide the empirical baseline necessary to design a sustainable dialogue framework for future curriculum integration.

Faculty: Giovanni De Francesco, CEEN

Email: gidefran@calpoly.edu

Accepted Projects Modes: In-person, Hybrid

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Wood houses dominate the construction landscape in the United States (U.S.), with over 94% of new homes constructed in 2022 being wood framed. New Zealand shares a similar construction landscape, with building practices for wood houses like those in California. Nevertheless, the 2010-2011 Canterbury earthquake sequence damaged nearly three quarters of the housing stock in the region. A total of 150,000 homes were damaged, with about one fifth exceeding NZ$100,000 in damage. To assess the likelihood of economic losses of California’s timber houses, a shake table testing campaign is underway at the California Polytechnic State University Advanced Technology Laboratory in San Luis Obispo. A full-scale, two-story timber specimen—designed with typical light-frame construction details—is to be tested with representative nonstructural components, including windows, doors, and sliding doors. This project will provide experimental evidence of earthquake-induced economic losses in timber houses and serve as a foundation for developing mitigation strategies to reduce the economic impact on our society of future earthquake events.

Faculty: Ann Goodell, CEEN

Email: anngood@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Approximately 11% of the U.S. construction industry is made up of women (Center for Construction Research and Training, 2025). Among women in the industry, 39% hold leadership roles, while 68% work in sales or office positions, underscoring the concentration of women in non-field roles. Participation has increased steadily year over year since 2010, per the U.S. Bureau of Labor Statistics (Gallagher, 2022). At the same time, construction is projected to grow faster than the national average, suggesting substantial opportunity for workforce entry and advancement. Despite these favorable conditions, women remain significantly underrepresented in the industry. Many women report experiencing sexual or racial harassment, discrimination in hiring, and gender bias on the job, which may contribute to retention and advancement barriers (Mefferd, 2024). The goal of this research project is to analyze the current available literature on women’s underrepresentation in construction and civil engineering and identify the primary factors most strongly correlated with this gender gap.

Faculty: Ann Goodell, CEEN

Email: anngood@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Reinforced concrete bridges experience progressive deterioration driven by chloride ingress and rebar corrosion, which reduces steel cross-section, weakens bond, induces cracking/spalling, and ultimately degrades stiffness and load-carrying capacity. This project develops and applies a finite element (FE) model that utilizes a simplified chemical-reaction/transport representation of corrosion initiation and propagation with structural response of reinforced concrete bridge components (e.g., decks and girders). Using a finite element program capable of incorporating corrosion-related material degradation, this study will simulate time-dependent deterioration under realistic exposure scenarios and quantify its effect on key performance metrics such as strength, stiffness, serviceability cracking, and dynamic effects. This finite element model will be compared with a real bridge under dynamic loads. The project will involve a close partnership with local jurisdictions to select and model a bridge which has corrosive damage. The work advances knowledge in bridge engineering by considering the effectiveness of the finite element mechanisms to capture true structural performance outcomes—supporting improved prediction of remaining capacity and better-informed maintenance and retrofit decisions.

Faculty: Kris Isaacson, CEEN

Email: krisaacs@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Hydrocarbon contaminants are frequently detected in drinking water supplies following urban wildfire and chemical spills. Yet despite common public guidance to use pitcher-style activated carbon filters as an interim mitigation measure, there are limited data quantifying their efficacy for these compounds under realistic use conditions. Preliminary results indicate that a representative pitcher filter exhibits incomplete benzene, toluene, ethylbenzene, and xylene (BTEX) removal, consistent with mass-transfer and contact-time limitations at typical operating flow rates. The overall goal of this work is to redesign the typical pitcher filter to maximize hydraulic retention time to better meet contaminant removal requirements in post-disaster situations. The SURP student will help lay the foundation for this project by quantifying the sorbent capacity for several commonly used pitcher-filters. Sorbent capacity will be determined for a selection of wildfire relevant organic compounds that span a range of chemical characteristics that impact sorption.

Faculty: Derek Manheim, CEEN

Email: dmanheim@calpoly.edu

Co-Advisor: Fatemeh Nayebloie (fnayeblo@calpoly.edu), CEEN

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Finished organic composts derived from dry thermophilic anaerobic digestion can concentrate microplastics (MPs), raising concerns about their mobilization into agricultural soils and subsequent risks to soil health, groundwater quality, and human exposure through food systems. MP release and retention are governed by complex, nonlinear interactions among abiotic factors such as ultraviolet irradiation, temperature, pH, and moisture, as well as soil properties that influence transport and chemisorption processes. However, mobilization pathways from compost to receiving soil microenvironments remain poorly characterized. This study proposes an integrated experimental and machine learning (ML) framework to elucidate and predict MP dynamics in compost-amended soils. Laboratory bench-scale 1-D flow-through column experiments will quantify MP leaching under controlled variations in polymer type, particle size (1 μm–5 mm), morphology, concentration, and environmental conditions. Experimental data will be used to train and evaluate ML models—including random forest, support vector regression, gradient boosting, and extreme gradient boosting—to identify dominant mobilization pathways and determine the relative importance of interacting abiotic drivers through an explainable AI approach. By coupling controlled experimentation with advanced data-driven modeling, this research aims to inform mitigation strategies, support sustainable compost management, and provide foundational data for future environmental and public health risk assessments.

Faculty: Stefan Talke, CEEN

Email: stalke@calpoly.edu

Co-Advisor: John Ridgely (jridgely@calpoly.edu), ME

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

In this SURP project, we propose to continue developing and implementing a prototype, radar-based water level sensor (~$100-200). Radar gauges have many advantages over other technologies used to measure coastal tides and sea-level rise. They measure the air-gap between the transducer and the water, and are therefore are less prone to marine degradation. Further, radar-measurements are (nearly) impervious to air temperature, humidity, wind, and other environmental fluctuations; for this reason, the National Ocean Service (NOAA-NOS) has recently been shifting from ultrasonic to radar technology at its tide gauges. However, radar technology requires significant power, which limits autonomous applicability, and commercial versions are prohibitively expensive ($5k-$25k). Recently, however, the price of radar transducers has reduced significantly, and we have successfully tested Cal-Poly developed radar gauges in Morro Bay for small time scale deployments (1-2 weeks). In this SURP, we will continue developing and testing this radar gauge. Significant effort is needed to optimize the software of a low-power microcontroller and integrate our design with solar panels capable of recharging on-board batteries. Additionally, we need to stress test our design under harsh marine conditions, and validate/quality assure measurements using time series from other instruments. Tests will be done to compare the radar sensor with other instrumentation, both in the lab and in the field. As needed, our pressure-gauge and DIY ultrasonic-based water level instrumentation will be refurbished and co-deployed along with the radar-based technology. Experience with programming, microcontrollers, electronic components, and instrumentation is preferred.

Faculty: Long Wang, CEEN

Email: lwang38@calpoly.edu

Co-Advisor: Jonathan Ventura (jventu09@calpoly.edu), CSSE

Accepted Projects Modes: Hybrid

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Microscopic imaging is essential to characterize multi-scale material behavior and understanding structure-property relationships. Recently our group has been developing a deep learning approach based on Generative Adversarial Networks (GANs) and Diffusion Models to reconstruct and artificially generate microstructures of strain-sensing nanomaterial networks based on microscope imagery. In this SURP project we would like to compare the performance of different models and evaluate alternative approaches to see if we can improve the quality of our results. Furthermore, we aim to test our approaches on a wider variety of materials, which will have different microstructures, to evaluate how versatile our models are.

Faculty: Long Wang, CEEN

Email: lwang38@calpoly.edu

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Structural Health Monitoring (SHM) is a data-intensive discipline for real-time assessment of structural integrity and safety. Modern SHM depends on applied computing (signal processing, ML/AI, and computer vision) to interpret sensing data and diagnose damage. This SURP project aims to develop and pilot a new course, CE 5558: Structural Health Monitoring, an interdisciplinary course that puts applied computing at the center of hands-on SHM investigations. The SURP student will build and test a low-cost SHM experimental kit, develop reusable MATLAB/Python notebooks and datasets for four course modules (sensing/data acquisition, damage feature extraction, ML-based damage detection, and computer vision), and help draft assessment artifacts to evaluate learning, teamwork, and ethical reasoning about data/algorithmic bias.

Faculty: Long Wang, CEEN

Email: lwang38@calpoly.edu

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Sensing technologies play significant roles in structural health monitoring (SHM) systems for monitoring and assessing structural conditions in real-time, which can enhance the safety and reliability of various structures. While engineered nanomaterial (ENM)-based sensors have remarkable potential to transform conventional sensing devices, large volume of ENMs released into the environment can significantly jeopardize the environment and public health. Thus, there is a pressing need to develop next-generation sensing materials in a more eco-friendly and sustainable manner. The goal of this interdisciplinary proposal is to sustainably develop sensing nanocomposites for monitoring structural damage by re-using waste materials as nano-/micro-scale functional material components. This project will lay the foundation for developing innovative, mechanically robust, and sensing structural materials in an eco-friendly, scalable, and low-cost manner.

Faculty: Ria Kanjilal, CPE

Email: rkanjila@calpoly@edu

Accepted Projects Modes: Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This research project aims to develop a robust, data-driven framework for non-invasive blood glucose prediction and anomaly detection using advanced machine learning (ML) and deep learning (DL) models applied to multimodal datasets. Traditional monitoring techniques are invasive, costly, and cumbersome, making them unsuitable for continuous, real-time tracking in free-living conditions. To address these limitations, this study integrates diverse physiological signals including heart rate, accelerometry, insulin records, dietary intake, and physical activity metrics to construct accurate and scalable predictive models for identifying glycemic fluctuations in real-time. The focus is on developing effective feature extraction techniques and predictive modeling approaches to enhance accuracy and reliability. By employing algorithms such as regression models, decision trees, LSTM networks, and Transformer-based architectures, this project aims to build models that generalize well across diverse subjects and conditions. These models will predict blood glucose levels and detect anomalies, including hypoglycemia and hyperglycemia, with high precision. The anticipated outcomes include creating accessible, non-invasive glucose monitoring systems capable of reliable performance in real-world conditions. Successful implementation will improve clinical decision-making, enhance patient outcomes, and offer a scalable, cost-effective alternative to traditional monitoring. This research has the potential to transform diabetes management by minimizing dependency on invasive techniques, improving the quality of life for individuals with diabetes, and expanding the applicability of machine learning in health monitoring systems.

Faculty: Ria Kanjilal, CPE

Email: rkanjila@calpoly@edu

Accepted Projects Modes: Hybrid

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project proposes a multimodal sensor fusion framework for emotion-aware Internet of Things (IoT) systems using publicly available multimodal datasets. Current emotion-aware IoT research largely relies on single-modality signals such as facial images or speech, making systems fragile to occlusion, environmental noise, sensor failure, and domain shift across users and contexts. Moreover, many existing approaches are evaluated in controlled laboratory settings and demonstrate limited robustness in real-world conditions. To address these limitations, this project develops a robust sensor fusion architecture that integrates inertial measurement unit (IMU) signals, facial imagery, and audio features for reliable affect recognition. IMU sensors capture posture, gesture, and movement dynamics; facial images encode expressive visual cues; and audio captures prosodic and spectral characteristics of speech. By fusing these complementary modalities through attention-based multimodal learning, the system aims to improve generalization, resilience to partial sensor dropout, and performance under noisy conditions. The project will compare classical machine learning baselines with deep multimodal architectures, including CNN-based visual encoders, temporal models for IMU signals, spectrogram-based audio networks, and adaptive fusion layers. Robustness will be evaluated through cross-subject validation and modality ablation studies. The anticipated outcome is a scalable, reproducible sensor fusion framework capable of supporting next-generation human-centered IoT systems.

Faculty: Sachiko Matsumoto, CPE

Email: sematsum@calpoly.edu

Accepted Projects Modes: In-person, Hybrid

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project will explore human-robot communication to support people’s situational awareness when working with robots. People controlling or collaborating with robots may have lower situational awareness for a variety of reasons, such as limited feedback from a robot or task constraints that require them to quickly switch their attention between several tasks. When people experience low situational awareness, it is not clear what communication modalities or strategies allow robots to support and continue collaborating with the person. This can be particularly relevant in telepresence applications or when the robot encounters a situation it cannot handle without human intervention. This project will explore context-appropriate communication modalities and strategies to improve robots’ abilities to collaborate with people.

Faculty: Sachiko Matsumoto, CPE

Email: sematsum@calpoly.edu

Co-Advisor: Andrea Schuman (anschuma@calpoly.edu), CPE

Accepted Projects Modes: Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project will integrate sociotechnical lessons into a Computer Engineering Human-Robot Interaction (HRI) course, not only in course meetings, but into labs, homework, and projects. The culture of engineering generally prioritizes neutral and apolitical engineering perspectives, which is associated with an isolated professional identity for students. Connecting technical engineering knowledge with broader contexts and motivations contributes to engineering identify formation. Thus, we will develop new material and revise existing material for an HRI course that will tie HRI concepts and methods to ethical and societal considerations, such as privacy, safety, ableism, and paternalism, among others. By demonstrating that sociotechnical knowledge is a fully integrated aspect of engineering, students will further learn how to create equitable engineering solutions. This content will be piloted and evaluated in a Fall 2026 offering of the course.

Faculty: Hisham Assal, CSSE

Email: hhassal@calpoly.edu

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

In this research we will investigate the current issues of deepfakes being submitted as evidence in the courtroom. As Artificial Intelligence becomes increasingly popular and people have more access to these technologies we need to be aware of deepfakes being submitted as evidence in a legal setting. We plan to train a model that will detect these images with a certain degree of accuracy. Furthermore, we plan to create a platform for law professionals to use in order to check photo evidence being submitted to a case. This will be a resource for judges, judicial support staff, attorneys, and litigants. We hope that this technology would eventually become a part of the legal process for verification and validation of images submitted as evidence in a court case.

Faculty: John Bellardo, CSSE

Email: bellardo@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Using the engineering design process the researcher will design and test a portion of the drag sail module by working to construct a prototype. The full drag sail module is intended to be flown on future CubeSat missions, demonstrating a deorbit within 5 years after end of life and reducing space junk from small satellites. This modular design is intended to reduce complexity when addressing CubeSat deorbit time. The project will give a student hands-on experience with industry-grade satellite subsystem design while also addressing the rapidly growing environmental and sustainability issue of space junk.

Faculty: Alexander Bisberg, CSSE

Email: abisberg@calpoly.edu

Accepted Projects Modes: In-person, Hybrid 

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Large-scale online games can offer spaces for identity exploration and social connection, but those benefits are not always distributed equitably across marginalized identities [5]. Sky: Children of the Light (Sky) is a social massive multiplayer online (MMO) game with unique design affordances to promote prosocial interactions. For example, unknown player avatars appear as gender-neutral silhouettes, which may shape how safe players feel about expressing their identity. Previous work on Sky have linked design elements like mentorship and generosity to downstream gains in social capital and reciprocity [1, 3, 4, 6, 7]. Existing analyses of gender in Sky have largely relied on binary categories (male/female), leaving gender-diverse players comparatively underexamined. This SURP project uses data already collected from an in-game survey, focusing on respondents who identify their gender as “Other.” The student will clean the survey data related to gender identity (e.g., missing values, fuzzy matching), query player survey responses, and link those responses to behavioral measures. The central research questions target whether Sky’s gender-neutral affordance functions as an identity-safety “buffer”: (1) Do players who identify as “other” gender feel less, equally, or more autonomous than those with identified with a specific gender? (2) Do players who identify as “other” gender receive less, the same, or more mentorship when first joining Sky?

Faculty: Dongfeng Fang, CSSE

Email: dofang@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Modern technology systems are rapidly integrating diverse technologies—such as AI services, cloud infrastructures, APIs, and wireless communications. These systems operate in increasingly adversarial environments that expand functionality but also but may unintentionally enable harmful or unanticipated misuse. In many cases, these risks are not failures of detection but consequences of design choices that did not fully anticipate adversarial use. As systems grow more interconnected, such misuse can scale faster than traditional reactive enforcement can respond. This proposal advances Anti-Abuse by Design, a preventive paradigm that embeds misuse resistance as a first-class architectural property. Rather than relying solely on reactive enforcement, we aim to embed adversary-aware access controls and built-in observability directly into system architecture to proactively constrain misuse. We will formalize anti-abuse design principles, develop measurable misuse-resilience metrics, and study systematic methods for identifying and characterizing misuse. Through an in-depth case study, we will evaluate how architectural interventions can reduce abuse while preserving usability, scalability, and trust.

Faculty: Daniel Frishberg, CSSE

Email: dfrishbe@calpoly.edu

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

We explore a class of algorithms known as Markov chain Monte Carlo (MCMC) algorithms, whose applications range from probabilistic modeling and document classification to statistical physics. The study of MCMC algorithms is a rich theoretical field. However, we take an empirical approach: we run the algorithms and attempt to evaluate their running time, more technically known as their mixing time. In part because the algorithms are randomized, evaluating the running time is not as easy as it sounds—but we have techniques for estimating it. The algorithm we will evaluate is a well-studied algorithm for randomly generating k-ary trees, a variation on the binary search trees many computer science majors encounter in a first-year data structures course. We use the same algorithm to randomly generate geometric tilings of a polygon with smaller polygons, known as k-angulations. The mentor advised a Cal Poly MS thesis in 2024 studying the special case of binary search trees (equivalently, triangulations). We will extend the code base from this project to the case of ternary search trees (equivalently, quadrangulations).

Faculty: Javier Gonzalez Sanchez, CSSE

Email: javiergs@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Human collaboration is shaped not only by explicit commands, but also by subtle perceptual cues such as gaze, gesture, posture, hesitation, and shared attention. These embodied signals enable people to coordinate fluidly, adapt to one another’s cognitive state, and build trust. In contrast, most human–robot interaction (HRI) systems remain primarily command-driven, relying on manual input devices and deterministic control models that lack awareness of the human partner’s perceptual and cognitive context. This project investigates how immersive presence and multimodal perceptual awareness influence operator performance and situational awareness in remote collaborative robotics. We aim to develop a platform that supports embodied collaboration between humans and robots. Building on prior work, the project will implement a remote interaction environment in which users control and monitor collaborative robotic arms and autonomous guided vehicles through an extended reality (XR) interface. The system will integrate brain–computer interface (BCI) signals, eye gaze tracking, gesture recognition, and body movement to improve inference of operator intent, cognitive workload, and shared attention. By fusing neural, visual, and motor signals, the platform will explore adaptive robot behaviors—such as pausing, confirming actions, adjusting pace, or signaling acknowledgment—in response to human cognitive and environmental cues. A real-time digital twin and cloud-synchronized communication layer will ensure low-latency interaction and safe operation. Expected outcomes include a modular research platform and empirical evaluation results that inform the design of next-generation perceptive and collaborative robotic systems.

Faculty: April Grow, CSSE

Email: amgrow@calpoly.edu

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

In collaboration with the department of Psychology and Child Development, we want to design and develop first-person 3D VR scenarios that confront the (teenage) player with positive but tough choices. It’s easy to say, in theory, that one might ask someone out, or take a club leadership role, but the increased immersion of a VR experience will more accurately test a player’s resolve! We currently have 6 hypothetical scenarios to model, including: a foreign language tour guide, taking a leadership role, trying a new sport, entering a sport competition, raising an objection, telling a joke, and asking someone new out. However, the depth, scope, number, and topics of scenarios are open to being redesigned with student input! We plan to develop the scenarios in Unity (C#) over SURP 2026, for a pilot study in Fall 2026 and a complete polished experience by Spring 2027. Experience with Unity, 3D modeling, and/or game design is a plus!

Faculty: Borislav Hristov, CSSE

Email: bhristov@calpoly.edu

Accepted Projects Modes: In-person, Hybrid

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Rapid advancements in next generation sequencing techniques has led to the development of a myriad of experimental techniques that allow researchers to measure different properties of single cells. However, these techniques are typically destructive, preventing scientists from measuring multiple properties at the same time, i.e paired measurements. Having disjoint datasets that reside in different spaces and that need to be integrated (paired) in order to build a fuller picture of the molecular processes in the cell is a key problem in computational multi-omics. Integrating these measurements to find a pairing between the cells is a particularly challenging task from the standpoint of computer science because the modalities may differ significantly in structure, scale, or noise characteristics, posing unique challenges for correspondence learning and classification. In this project, we aim to develop a novel deep learning framework for finding a shared latent embedding of the cell measurements. The key idea is that two measurements that come from the same cell should be close to one another when projected in the latent space. We will use a small subset of the data as training cells to learn separate autoencoder models per modality and then employ cross training (or GAN) to align their embedding spaces. Once trained, this dual autoencoder architecture can be used to project a cell from one measurement space into another by using the encoder1 and then decoder2. The ultimate goal is to have a deep learning model that is capable of efficiently integrating large scale multi-omics datasets.

Faculty: Irene Humer, CSSE

Email: ihumer@calpoly.edu

Accepted Projects Modes: Fully Remote, Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

The Poisson equation is a fundamental partial differential equation that arises in many areas of science and engineering, including heat diffusion, electrostatics, fluid flow, and image processing. Discretization of the Poisson equation leads to large, sparse linear systems whose efficient solution is critical for practical simulations. In this project, the student will study and implement several classical numerical methods for solving Poisson-type linear systems, including Jacobi iteration, Gauss–Seidel, the Conjugate Gradient (CG) method, and multigrid techniques. The project will emphasize how matrix structure such as sparsity, symmetry, and positive definiteness affects solver performance and scalability. Through computational experiments in one and two spatial dimensions, the student will compare convergence behavior, runtime scaling, and robustness of these methods. Depending on project progress, the student may also explore special-case solvers, such as FFT-based methods for periodic boundary conditions, and examine opportunities for parallelism in iterative algorithms. The primary outcomes will be reproducible numerical experiments, visualizations of solver convergence, and a research poster suitable for presentation at the SURP symposium and potential submission to an undergraduate research or educational venue.

Faculty: Paris Kalathas, CSSE

Email: pkalatha@calpoly.edu

Co-Advisors: Franz Kurfess (fkurfess@calpoly.edu), CSSE and Jenny Zheng Wang (jwang96@calpoly.edu), IME

Accepted Projects Modes: Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This proposal describes a research activity aimed at designing and evaluating prompt scaffolding strategies for agentic tutoring environments powered by large language models (LLMs). As LLMs and LLM-driven tutoring systems become parts of the educational settings, it is critical to understand how carefully structured prompts can guide learners through complex problem-solving tasks promoting the ethical use of AI. The goal of this research is to design and test prompt scaffolding strategies that can be integrated into agentic learning environments for CS laboratories. Historically, scaffolding theory has informed instructional design by providing temporary support structures (Wood, Bruner & Ross, 1976). The emergence of LLM-based systems creates a new infrastructure to operationalize such scaffolding dynamically, adapting in real time based on learner’s responses and errors (Kasneci et al., 2023). Research in prompt engineering has demonstrated that the structure, specificity, and sequencing of prompts significantly influence the quality and pedagogical value of AI-generated responses (White et al., 2023). This project extends that work into an educational context, asking how prompt design can be harnessed intentionally to support learning rather than merely to elicit correct answers.

Faculty: Fahim Khan, CSSE

Email: fkhan19@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project is a continuation of my prior SURP 2025 project that developed a mobile app and created an annotated dataset to train computer vision models for detecting infrastructure relevant to accessibility in public spaces (e.g., curb ramps, crosswalk features, sidewalk conditions, and obstacles). Building on that foundation, the follow-up project will extend the system from smartphone-based capture to wearable and edge-AI deployment using AI camera glasses and NVIDIA Jetson Orin Nano boards. The goal is to enable real-time, hands-free sensing and feedback, while also expanding data diversity and improving model robustness in varied environments (lighting, weather, surface types, and urban layouts). Students will be onboarded with structured training so they can effectively build on top of the existing codebase, dataset, and model pipeline from the previous project. The team will integrate the camera-glasses streaming pipeline with the Jetson Orin Nano for on-device inference, optimize models for low-latency performance, and connect detections to the existing mapping/reporting workflow. The final deliverables will include a wearable prototype system, updated datasets and models, evaluation results, and a deployment-ready demonstration in real public settings.

Faculty: Franz Kurfess, CSSE

Email: fkurfess@calpoly.edu

Co-Advisor: Jenny Wang (jwang96@calpoly.edu), IME and Paris Kalathas (pkalatha@calpoly.edu), CSSE

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project investigates how structured ethical rules influence agent behavior within collaborative AI environments. As part of foundational research in the context of Agentic Design frameworks, this study focuses on designing and evaluating an ethics-monitoring layer to mitigate undesirable behaviors such as hallucinations or unsafe decisions in multi-agent systems. Using frameworks like AutoGen, the student will implement a specialized “constraint-monitor agent” capable of performing real-time privacy and safety checks. The research aims to determine if explicit constraint layers can strengthen system reliability and provide a robust foundation for responsible AI in complex domains. The work is directly relevant to several other ongoing activities like AI for Search and Rescue, Agentic Design for Engineering Labs, and Sustainability in AI.

Faculty: Joydeep Mukherjee, CSSE

Email: jmukherj@calpoly.edu

Accepted Projects Modes: Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project proposes the design and implementation of a small-scale Internet of Things (IoT) system that allows users to remotely display text messages on an LED display using a mobile phone connected to any cellular network. The system will integrate an Arduino microcontroller, a Raspberry Pi, and an LED matrix or LED panel, coordinated through a mobile-friendly web application. Messages sent from a smartphone will be transmitted over a cellular data network to the Raspberry Pi, which will process the request and relay display commands to the Arduino for real-time output on the LED. From a technical perspective, students will work with embedded systems programming, device-to-device communication, and networked application design. The Raspberry Pi will function as a lightweight server, handling incoming HTTP or MQTT requests from the mobile application, while the Arduino will manage low-level control of the LED hardware. Students will explore serial communication, basic networking protocols, and message parsing, as well as system reliability and latency considerations in remote device control. The mobile interface will be implemented as a simple web or hybrid application, enabling cross-platform access without reliance on a specific carrier or operating system. This design emphasizes interoperability and real-world IoT constraints, such as connectivity, scalability, and security. The project will be delivered as a hands-on learning module in which students build, program, test, and refine a complete IoT system. By integrating hardware, software, and networking components, the project provides a comprehensive experiential learning opportunity that prepares students for advanced work in IoT development, mobile applications, and networked systems engineering.

Faculty: Sumona Mukhopadhyay, CSSE

Email: mukhopad@calpoly.edu

Accepted Projects Modes: In-person, Hybrid

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project aims to develop a Python-based interactive visualization platform for analyzing EEG brain activity in relation to cognitively meaningful events in learning tasks. Inspired by professional neurophysiology tools such as MNE, the proposed system extends traditional EEG viewers by integrating time-locked experimental events (e.g., question onset/offset, auditory markers), survey text, and participant responses into a unified interface. The platform will enable researchers to visually inspect raw and preprocessed EEG signals alongside protocol-defined events and associated textual stimuli, supporting exploratory analysis, data quality assurance, and interpretability of multimodal learning data. Designed entirely in Python, CANDY emphasizes reproducibility, modularity, and extensibility, and serves as a foundation for future machine learning–based cognitive event modeling and human-in-the-loop annotation workflows.

Faculty: Dev Sisodia, CSSE

Email: dsisodia@calpoly.edu

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Wildfires increasingly threaten California communities, yet most existing risk maps operate at coarse regional scales and rely on opaque models that fail to link risk scores to clear, verifiable mitigation actions. Building on two years of prior work by the Perch Sensing team, including structured interviews with homeowners, fire officials, insurers, and researchers, this project addresses a key gap: mitigation efforts are often performed but remain undocumented or unverifiable, contributing to mistrust and insurance instability. The team previously developed a conceptual wildfire framework organized around three components – Predict, Prevent, and Protect – capturing ignition potential, structural vulnerability, vegetation characteristics, topography, and community factors. This SURP project will operationalize that framework by developing an interpretable, property-scale wildfire risk model using publicly available geospatial data. The student researcher will implement transparent statistical methods to decompose risk into defensible-space and exposure components, incorporate uncertainty quantification, and simulate how mitigation actions change predicted risk. Outcomes will include a reproducible modeling pipeline, technical report, and SURP poster, advancing research at the intersection of computing, fire science, and community resilience.

Faculty: Jonathan Ventura, CSSE

Email: jventu09@calpoly.edu

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

3D reconstruction methods such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) achieve photorealistic view synthesis by reconstructing a detailed model of a scene from a collection of input images. However, such models are unable to model missing areas of the scene, such as occluded regions. We propose to incorporate multi-view image generation models into the scene optimization process, to combine data-driven scene reconstruction with generative scene completion. Ideal candidates will have experience in computer vision, 3D graphics, and deep learning.

Faculty: Austin Wright, CSSE

Email: awrigh20@calpoly.edu

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Bayesian graphic models are frequently used in many forms of data analysis, in particular in cases of causal inference. Their visual structure makes them supposedly much more amenable to rapid inspection when compared to explicitly written out joint and conditional probability distribution factorizations. However, there does not currently exist work explicitly studying the effect of graphical representations compared to symbolic representations on the process of data analysis as done by humans. This project will first develop in interactive visualization interface for creating and applying Bayesian networks to data analysis and visualization, and then systematically compare the efficacy of graphical and interactive visualization-based methods compared to existing code-based methods with a user study. While it is so frequently taken for granted that visual representations and diagrams make understanding complex problems easier, this is rarely studied precisely. Finally, developing the interactive tool for the study, and releasing it freely in open-source, will provide a valuable resource not only for people doing analysis with graphical models, but also for educators teaching such methods.

Faculty: Anu Aggarwal, EE

Email: aaggar24@calpoly.edu

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Goal: of this project is to analyze animal data from the Moser lab to decipher connectivity in the brain’s spatial navigation system. Background: Place cells [1] and grid cells [2] are the neurons that help animals remember the spatial and time information in a scene from episodic memories. Place cells are located in the hippocampus and grid cells are located in the entorhinal cortex of the brain. Both the cells are known to produce firing patterns that are highly correlated with animal position in space. The grid cells are the only inputs to the place cells. However, it is not entirely clear how the place cells process information received from grid cells. To understand this, it is important to understand their connectivity to grid cells. Hence, in this project, we took data from the Moser lab. The data includes single cell recordings from 342 CA3 neurons in 8 animals in 11 different rooms. Of these 342 cells, only 2 fired in all the 11 rooms and over 100 fired in one room. Methods: To understand connectivity between the grid and the place cells, we will create grid cell firing patterns in Matlab. Connection weights between the place and grid cells will be learned using machine learning- gradient descent algorithm. Once correct place cell firing patterns are learned, we will study how many and what kind of connections can replicate them. We have already presented our initial results from this project at Society for Neuroscience Conferences [3][4]. It is a work in progress, and we need an undergraduate student to complete the simulations. Outcome: this will help us understand the connectivity between the grid and place cells and understand their function. This work will be presented in international research journals like Neuroscience and conferences like the society for neuroscience (SFN).

Faculty: Dennis Derickson, EE

Email: ddericks@calpoly.edu

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

The 122 GHz to 123 GHz frequency band has been allocated for Industrial, Scientific, and Medical (ISM) applications and is also allocated to Amateur Radio as a secondary user by the IEEE and the International Telecommunications Union (ITU), https://www.itu.int/en/Pages/default.aspx . This ISM band is currently experimental in nature due to the lack of commercial off-the-shelf equipment availability. This frequency band has high propagation loss due to atmospheric absorption. Atmospheric absorption makes it a poor choice for long-distance communication. Conversely atmospheric absorption and the 1 GHz frequency width make it a great candidate for high-bandwidth, short distance, connections that can be easily re-used at nearby nodes without interference. The project goal is to have a SURP student design a 122.25 to 123 GHz frequency downconverter for this new WiFi band using the United Monolithic Semiconductor (UMS) (https://www.ums-rf.com/) diode process. This wifi band has a lot of future utility as high frequency semiconductor processes are developing to the point where the designs are now becoming feasible. High frequency integrated circuit design skills are sought after by both defense contractors and the commercial communication employer sectors. The SURP student’s project goal is to develop a custom radio frequency integrated circuit and document the process steps so that other students can easily replicate the work. The SURP student should leave the project with proficiency using the Keysight ADS computer aided design tool. The SURP student can do this work at any location.

Faculty: Dale Dolan, EE

Email: dsdolan@calpoly.edu

Accepted Projects Modes: Hybrid, In-person, Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

The Cal Poly Solar farm has been built as a single axis tracking facility with two different types of panels. Both conventional single cell solar panels and twin cell solar cells have been used in its construction. Twin Cell panels typically perform better than their conventional counterparts when shaded by other panels in the row in front of them in a fixed tilt system. However neither module performs well when even a small portion is shaded. This project will access the system API to access data for the field and process to determine the energy yield improvements that can be made by optimizing backtracking settings for each array in the Cal Poly Solar Farm. Backtracking is a technique where the single axis tracker avoids shading other rows by sub-optimally tracking the sun until it is high enough in the sky that this is not necessary. This is relatively straightforward in a field that is flat or at least has all controlled modules in the same array. However the Cal Poly Solar farm installation has significant variation in terrain and this was transferred to the modules in adjacent rows such that they are not in the same plane nor are they parallel. As a result there is significant error in the backtracking algorithm that was not designed for this case which is seen by significant power losses in both of the types of panels. The subject of this research project is to determine the improvements that can be made in energy yield by adjustment of the backtracking settings for each array. The parameters have been adjusted and we will need to determine the change in daily energy yield that is achieved for all 12 trackers. This is extremely time intensive and we will be using automation to implement the calculations and comparisons, likely through use of Python. I have obtained access to allow data acquisition for power and energy data and operation of the tracking system at the Cal Poly Goldtree Solar Farm to allow this project to be completed 100% remotely if necessary. All control and data collection can be accomplished via a web based dashboard that allows us to monitor the performance of the farm in real time and to collect data that can be analyzed to determine the improvements.

Faculty: Siavash Farzan, EE

Email: sfarzan@calpoly.edu

Accepted Projects Modes: Hybrid, Fully Remote, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project develops and validates an end-to-end engineering pipeline for mini quadrotor autonomy, centered on repeatable trajectory tracking across simulation and hardware. A physics-based digital twin is used to develop and stress-test estimation and control components under realistic sensing, lighting, and disturbance conditions, then verify transfer to a physical mini quadrotor with minimal retuning. During Summer 2026, the SURP student will close remaining gaps needed for a clean, reproducible release: (1) finalize simulation environments and automated regression tests, (2) integrate a complete logging and evaluation harness for tracking benchmarks, (3) run robustness sweeps over key uncertainty sources (e.g., mass, motor dynamics, sensor bias, latency, and environmental disturbances), and (4) execute a focused sim-to-real validation campaign to quantify performance on representative tracking tasks. The outcome is a documented engineering artifact (code, configuration, and benchmark results) that enables rapid iteration on aerial autonomy while reducing reliance on dedicated flight facilities.

Faculty: Kun Hua, EE

Email: kuhua@calpoly.edu

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This research project investigates the feasibility, reliability, and predictive performance of wearable biosensing technologies for real-time driver state monitoring. The study focuses on wearable electrocardiography (ECG), photoplethysmography (PPG), galvanic skin response (GSR), and motion sensing (MOT) to detect stress, cognitive workload, and fatigue during simulated driving tasks by using the existing Shimmer healthcare development kit, iPhone and iWatch. The project advances knowledge in biomedical signal processing, machine learning, and intelligent transportation systems by addressing the challenge of distinguishing stress-specific physiological signatures from general autonomic arousal in dynamic environments. By integrating robust signal processing with adaptive classification models, this work bridges electrical engineering, biomedical engineering, and human factors research. The outcomes will contribute to safer intelligent transportation systems and improve wearable sensor reliability under motion conditions.

Faculty: Jenna Kloosterman, EE

Email: jklooste@calpoly.edu

Co-Advisor: Stephen Broccardo (stephen.p.broccardo@nasa.gov), Bay Area Environmental Research Institute, NASA AMES

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Atmospheric science would like to observe molecular species in Earth’s atmosphere in the Sun’s absorption spectra via a mesospheric balloon platform. The instrument on the balloon must be able to track the sun and have stable pointing as it ascends through the atmosphere. The current tracking and stabilization method predicts and corrects changes in actuator speed and angular velocity. The addition of an angular acceleration sensor would add a significant upgrade because it would enable the system to predict and correct for actuator torque. No such component exists that fits the needs of our system. This project will investigate two different approaches for building an acceleration sensor. The first (1st) approach utilizes commercial-off-the-shelf (COTS) linear accelerometers in a circle. The second (2nd) approach uses two liquid filled spiral channels in opposite directions with a differential pressure sensor. The student will be responsible for researching each approach and deciding on the best approach based on analysis and cost in collaboration with the advisor. After deciding on the approach, the student will build and test the chosen sensor. The student will also be responsible for working with a second student who will implement a control loop to stabilize instrument. Together they will test and demonstrate the stabilization system.

Faculty: Jenna Kloosterman and Dennis Derickson, EE

Email: jklooste@calpoly.edu,  ddericks@calpoly.edu

Industry Partner: Hascall-Denke

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Faculty: Souvik Kundu, EE

Email: sokundu@calpoly.edu

Accepted Projects Modes: Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Biosensing technologies, such continuous glucose monitors (CGMs) and wearable diagnostic gadgets, have changed how we treat chronic diseases. But in the United States, access to these technologies is still quite unequal based on race and income. Structural inequities in access are caused by high out-of-pocket prices, differences in insurance coverage, regulatory processes, and market concentration. If people don’t do anything on purpose, new technologies could make health inequities worse instead of better. This research uses the Design Justice framework (Costanza-Chock, Design Justice, MIT Press, 2020: An exploration of how design might be led by marginalized communities, dismantle structural inequality, and advance collective liberation and ecological survival) to look at new biosensing technologies. Design Justice contends that design processes frequently perpetuate structural disparities when they prioritize profit, institutional convenience, or dominant demographics above those most affected by systemic barriers. Instead, design needs to take power, access, and material circumstances into account directly. Based on this foundation, the project employs datasets that are harmonized and represent the whole country: • IPUMS MEPS (Medical Expenditure Panel Survey) https://meps.ipums.org/meps/ • IPUMS NHIS (National Health Interview Survey) https://nhis.ipums.org/nhis/ These databases contain labeled, individual-level information on healthcare costs, insurance coverage, the prevalence of chronic diseases, delayed care owing to cost, and demographic factors. Two undergraduate students is proposed to work together in this project**. **However, if funding is not available for 2 students, the scope of the project can also be reduced to fit 1 student for 8 weeks duration. In that case, I would request the review committee to please evaluate the proposal w.r.t. student 1 or student 2 only. Student 1 will use IPUMS MEPS and NHIS to provide an empirical analysis of differences in healthcare spending, out-of-pocket costs, insurance coverage, and avoiding care because of costs based on race and income. Student 2 will perform a structured, justice-informed analysis of technological and socio-economic strategies to enhance biosensor affordability, focusing on material selection, simplification of device architecture, scalability of manufacturing, reimbursement frameworks, intellectual property arrangements, and market concentration trends. This initiative prioritizes affordability as a justice-oriented design variable, creating a framework for equity-informed biosensing innovation, rather than viewing it as a secondary restriction.

Faculty: Souvik Kundu, EE

Email: sokundu@calpoly.edu

Accepted Projects Modes: Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Diabetic retinopathy is one of the most common causes of blindness that can be avoided. Accurate identification of fine retinal blood vessels in fundus pictures is essential for early diagnosis. Yet, these images frequently exhibit low contrast, uneven lighting, and noise that mask minute vascular structures. This project examines the efficacy of quantum-inspired image processing strategies, executed through classical quantum simulation frameworks, in enhancing retinal vessels relative to conventional image processing methods. The student will use the DRIVE (Digital Retinal Images for Vessel Extraction) dataset, which is open to the public (Dataset link: https://drive.grand-challenge.org/) The project will look at classical methods of enhancing images (like Fourier filtering, wavelet transforms, and contrast enhancement) and quantum-inspired methods (like amplitude encoding-based image representations, quantum Fourier transform (QFT)-inspired filtering, and hybrid quantum-classical pipelines simulated in Python (like Qiskit or something similar). The objective is to assess if quantum-inspired computational architectures provide enhanced contrast, edge retention, or feature differentiation for intricate vascular structures pertinent to diabetic retinopathy screening. In the wake of the AI revolution, attention is shifting toward quantum computing as a potentially transformative computational paradigm. By integrating quantum-inspired techniques into biomedical image analysis, this project equips students with foundational skills in an emerging domain poised to influence the future of sensing and diagnostics.

Faculty: Jason Poon, EE

Email: jasonp@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project investigates the use of analog computing to accelerate power system simulations for grid interconnection studies. The research will focus on designing and evaluating analog computing circuits capable of efficiently modeling large-scale power networks with detailed representations of power electronics–dominated loads. A software interface will be developed to allow users to integrate the analog hardware with standard desktop-based simulation tools. Through circuit design, simulation, fabrication, and experimental validation, the project aims to demonstrate how analog computing can significantly reduce computation time while maintaining sufficient accuracy for practical interconnection analysis.

Faculty: Puneet Agarwal, IME

Email: pagarw05@calpoly.edu

Accepted Projects Modes: Hybrid, Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

In an age of constant exposure to digital information, distinguishing credible content from factual inaccuracies has become a challenge. A growing accessibility to generative AI tools has amplified the already prevalent spread of misinformation on social media, and yet potential use of these same systems to mitigate misleading narratives remains largely unexplored. This research project seeks to evaluate the ability of advanced large language models (LLMs) to detect and assess the validity of information across diverse forms of text media. Building on an established proof of concept from our 2025 summer pilot, we now aim to deepen our understanding by implementing nuanced scoring systems and condition-specific datasets that will produce findings on when and how LLMs reliably detect various textual forms of misinformation. We will construct several datasets containing both factual and misleading content, including human-authored news articles from both reliable and unreliable sources, professionally fact-checked short-form claims, AI-generated long-form articles with defined agendas, and AI-rewritten versions of human-authored texts. Each item will be thoroughly reviewed and labeled with ground scores to ensure reproducibility. We will prompt three state-of-the-art LLMs (GPT-5, Claude, and Grok) to evaluate each sample across four metrics: factual accuracy, intent to deceive, emotional manipulation, and confidence in their evaluation, to produce a multi-dimensional analysis of both class and severity of misleading content. Each model will return structured scores using a standardized prompt framework to ensure consistency. Ground-truth labels are derived from verified fact-checking sources and structured human evaluations, while cross-model consensus will be measured as an independent reliability metric. Our analysis seeks to clarify the limitations of LLMs in real-world misinformation detection and uncover the conditions under which the LLMs perform best. This project is set to produce three deliverables; first, transparently labeled datasets with associated ground-scores, which will be made publicly available to advance misinformation research; second, a peer-reviewed publication documenting our methodology and findings; and third, a web application that allows real-time evaluation across our four metrics, enabling users to input text and receive immediate feedback. We aim to build upon the established benchmarks for AI-assisted information evaluation, providing researchers and consumers with functional tools to navigate the modern information landscape. By prioritizing reproducibility and transparency, we will ensure findings are both verifiable and ethically communicated.

Faculty: Duha Ali, IME

Email: duali@calpoly.edu

Co-Advisors: Rafael Silva (rguerras@calpoly.edu), ITP and Javier Sanchez (javiergs@calpoly.edu), CSSE

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Short-form social media platforms (e.g., TikTok, Instagram Reels, YouTube Shorts) promote rapid content switching, high novelty, and brief engagement cycles. Emerging literature suggests that heavy exposure to such environments may be associated with increased distractibility, higher switching tendencies, and reduced sustained attentional control. However, it remains unknown whether these attentional patterns translate into reduced persistence during goal-directed interaction with conversational AI systems. This project investigates whether habitual and immediate short-form media exposure influence “attention tolerance” during AI interaction, defined as the ability to sustain engagement with a single AI-mediated task long enough to reach a high-quality outcome. Using a mixed design that combines behavioral measures, AI interaction telemetry, and EEG-based engagement metrics, we will recruit 24 adult participants (ages 18–35) for controlled lab sessions. The SURP student will assist in participant recruitment, data collection, AI session instrumentation, preprocessing of EEG data, and preliminary statistical analysis. Results will advance understanding of how modern media environments shape cognitive engagement in AI-assisted work contexts.

Faculty: Mohamed Awwad, IME

Email: mawwad@calpoly.edu

Accepted Projects Modes: Hybrid, In-person, Fully Remote

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Large Language Models (LLMs) are increasingly proposed as decision agents in supply chain planning, procurement, and logistics systems. However, existing research primarily demonstrates single model implementations without systematic cross-model benchmarking, standardized evaluation protocols, or quantitative validation across operational scenarios. This project develops a structured and reproducible benchmarking framework to evaluate multiple large language models on defined supply chain decision tasks such as inventory control, disruption response, and supplier selection. Using consistent input conditions and scenario variations, models will be assessed against operational baselines using measurable performance metrics, including cost impact, service level effects, and output consistency. By providing empirical comparison across models within a controlled decision environment, the project advances methodological rigor in evaluating generative AI for industrial engineering applications and contributes new evidence on the practical performance and limitations of LLM-based decision support systems in supply chain contexts.

Faculty: Aditya Chivate, IME

Email: achivate@calpoly.edu

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Modern manufacturing increasingly relies on robotic systems, yet many automated processes still require human oversight when uncertainty arises. Digital twins (DT) virtually mirror physical systems and are often used to monitor robots. Most DTs do not account for how humans actually interact with these systems in practice. This project explores human-in-the-loop (HITL) digital twin approach for robotic assembly, where a virtual model predicts task success and requests human input only when the confidence is low. This project emphasizes practical skills in robotics system integration, simulation, data collection, and human–robot interaction. The focus is on building a working system and gaining experience with real hardware, rather than developing complex algorithms. The outcomes will prepare the student for careers in robotics, automation, or software engineering, while also laying the groundwork for future research in human-centered digital twins.

Faculty: Byeongmok Kim, IME

Email: bkim142@calpoly.edu

Accepted Projects Modes: In-person, Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project proposes a collaborative intelligence framework to support human decision-making in multi-robot systems under multi-objective tradeoffs. In complex human–robot collaboration scenarios, Pareto optimization can generate multiple optimal coordination strategies balancing objectives such as efficiency, safety, energy consumption, and workload distribution. However, selecting among these alternatives can be cognitively demanding for human operators. The proposed framework introduces a generative decision-guidance module that translates Pareto-optimal solutions into structured, human-readable tradeoff summaries, interactively incorporates operator preferences, and provides explainable, data-driven justifications derived from multi-robot system states. The system will be implemented and evaluated within a computer simulation environment modeling multi-robot collaborative tasks. Performance will be assessed in terms of decision quality, interpretability, and decision efficiency. The expected outcome is a validated human-centered decision-support framework for collaborative robotics.

Faculty: Jill Speece, IME

Email: jespeece@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

TurboRad’s radiology reporting software is integrating artificial intelligence to support radiologists during report creation. The AI system is designed to offer real-time diagnostic suggestions that help ensure all clinically relevant conditions are considered, based on the reason for exam, the radiologist’s preliminary findings, and specific patient demographics. This summer research project will recruit two undergraduate students to help expand this AI-driven capability and to verify and validate its effectiveness in real-world reporting scenarios. Students will contribute to refining the logic behind diagnostic suggestions, evaluating system performance across diverse clinical cases, and assessing how well the AI supports thorough and accurate radiology reporting without disrupting clinical workflow.

Faculty: Jenny Wang, IME

Email: jwang96@calpoly.edu

Co-Advisors: Prof. Franz Kurfess (fkurfess@calpoly.edu), CSSE and Paris Kalathas (pkalatha@calpoly.edu), CSSE 

Accepted Projects Modes: Hybrid

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Collaborative multi-agent AI systems increasingly distribute tasks across specialized agents to solve complex problems. However, when agents produce contradictory recommendations, pursue competing objectives, or operate with incomplete information, system-level conflicts can emerge. This project investigates structured methods for detecting and resolving conflicts in multi-agent systems through experimentally evaluated arbitration mechanisms. The student will design controlled conflict scenarios and implement alternative resolution strategies—including hierarchical override, voting-based arbitration, and mediator agents—to evaluate their impact on system stability, decision quality, and computational efficiency. The project advances knowledge in agentic AI architecture by developing empirically grounded design guidelines for robust multi-agent coordination.

Faculty: Xuan Wang, IME

Email: xwang12@calpoly.edu

Co-Advisor: Jenny Wang (jwang96@calpoly.edu), IME

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Adolescent Idiopathic Scoliosis (AIS) is a three-dimensional spinal deformity for which brace design remains largely empirical, with limited mechanistic understanding of brace–torso–spine interaction. This project aims to develop and biomechanically validate a patient-specific digital twin framework that integrates radiation-free 3D torso surface scanning with radiographic spinal alignment to simulate brace-induced mechanical loading using finite element methods. By constructing subject-specific computational models, the research will quantitatively characterize corrective force transmission, pressure distribution, and spinal deformation response. The work advances knowledge in computational biomechanics by establishing a validated, patient-specific modeling approach for evidence-based brace optimization.

Faculty: Zhiyuan Wei, IME

Email: zwei03@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Emergency services, such as fire and medical response, are essential for safeguarding lives and ensuring public safety during crises. However, current practices of first responders in fire departments rely heavily on experience, which may limit the effectiveness and adaptability of decision-making in dynamic and uncertain environments. This project aims to develop a data-driven contextual stochastic optimization (CSO) framework to enhance fire department service planning and real-time operational decision-making under uncertainty. First, multi-source data will be leveraged to extract and quantify contextual information characterizing operational uncertainty. Next, we develop an end-to-end learning and optimization framework that integrates predictive models and operational decisions. This framework captures the stochastic nature of fire engine availability using contextual information to dynamically adapt deployment decisions to evolving incident conditions. Lastly, computational experiments will be conducted to demonstrate that the proposed end-to-end CSO framework enhances overall system resilience and robustness against uncertainty. This project provides students with hands-on experience in applying data analytics and stochastic optimization modeling techniques to real-world challenges in emergency response management. The insights generated through this work have the potential to support a shift from reactive, experience-based practices to proactive, data-informed operations, contributing to more resilient and responsive emergency service systems.

Faculty: Nicole Johnson-Glauch, MATE

Email: njohns66@calpoly.edu

Accepted Projects Modes: In-person, Fully Remote, Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Have you ever wondered whether the classes you take and the order in which you take them are setting you up to be a successful engineer? As a state institution, Cal Poly’s primary goal is to serve all students of California with high quality, Learn by Doing educational experiences. Could our engineering curriculum be serving some students but not others? This project idea arose from noticing that the six-year graduation rate differs across different engineering programs, especially in the General Engineering (GENE) program. The six-year graduation rate for GENE ranges from 14-62 percentage points lower than the six-year College of Engineering graduation rate for cohorts who started in Fall 2013 – Fall 2019. This is surprising because GENE has the most flexible program with 40 units dedicated to technical electives to customize the degree. The College of Engineering is interested in understanding how systemic structures such as program complexity impact the lived experiences of engineering students from different backgrounds across the college of engineering, especially in GENE. A student on this project could explore the relationship between program complexity, students’ socio, economic, or demographic backgrounds, and student success metrics (time to graduation, job placement, etc…). Alternatively, a student could explore whether the intended progression through engineering curricula is the one most commonly experienced by students from different backgrounds. Results from this study will be used to inform how we could revise engineering curricula, develop tools to more effectively advise engineering students, and identify areas where college-level reform could better support students.

Faculty: Nahmoon Kim, MATE

Email: nkim113@calpoly.edu

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Cyclic oxidation resistance plays a critical role in the durability of high-temperature turbine alloys, as insufficient resistance can lead to oxide spallation, material loss, and reduced component lifetime in energy and aerospace systems. In the 1990s, NASA conducted extensive cyclic oxidation experiments across a wide range of alloy compositions and temperatures, generating a valuable dataset that can benefit from modern quantitative analysis. This project will use this legacy NASA dataset to investigate how alloy composition influences cyclic oxidation resistance through a structured, data-driven approach. The student researcher will organize compositional and oxidation data into a reproducible format, extract key oxidation metrics, and analyze compositional trends across alloys and temperatures. By combining classical oxidation metallurgy with modern data-analysis tools, this work advances understanding of composition–property relationships in turbine alloys and enables more systematic comparison of oxidation performance across complex alloy systems.

Faculty: Kareesa Kron, MATE

Email: kkron@calpoly.edu

Co-Advisor: Jennifer Mott (jpeuker@calpoly.edu), ME

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Cal Poly, San Luis Obispo is interested in improving the accessibility of course materials and improving the perceptions of accessibility tools amongst faculty members. Dr. Mott and Dr. Kron have been working towards this goal, including outreach to faculty and work on accessibility tools in the context of designing course materials. A survey is currently being collected at department meetings that aims to assess current perceptions of and implementation of accessibility in course material design by Cal Poly faculty. This SURP project will start by processing and analyzing the data of the survey(s) collected. The data analysis will enable us to describe current faculty perceptions and predict potential areas where faculty perceptions are poor and should be targeted with further training. The second portion of the project will prepare surveys and assessments that will be deployed in MATE 210 to explore the impact of alternative grading on students with accommodations. Alternative grading is frequently described as more accessible, but it is under-researched as to how it affects students with varying accommodation needs.

Faculty: Kareesa Kron, MATE

Email: kkron@calpoly.edu

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project aims to develop high-quality polyurethane (PU) foam for surfboard cores using recycled polyethylene terephthalate (PET) as a primary feedstock. Post-consumer PET will be chemically depolymerized through glycolytic transesterification to produce hydroxyl-terminated polyester oligomers. These recycled polyols will then be reacted with castor oil and methylene diphenyl diisocyanate (MDI) to form a polyurethane prepolymer, which will be foamed through controlled CO₂ generation. The resulting foam will be evaluated for density, buoyancy, water absorption, and UV stability to determine its suitability for marine environments. The goal is to create a structurally sound, lightweight, and environmentally sustainable alternative to petroleum-based polyurethane foam.

Faculty: Brendon Anderson, ME

Email: bga@calpoly.edu

Accepted Projects Modes: Hybrid, Fully Remote, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

The design and analysis of modern engineering systems often amounts to solving high-dimensional nonconvex optimization problems with potentially many local-but-not-global optima. Due to getting “stuck” at such local optima, conventional gradient-based optimization algorithms result in estimated losses on the order of billions of dollars per year in large-scale industries such as power grid management and artificial intelligence (AI). In this project, we will design, implement, and evaluate new optimization methods with the explicit aim of solving for global optima. Specifically, we will explore techniques to recast engineering optimization problems into game-theoretic models of large populations of agents playing a strategic game. Social and behavioral intuitions of the models will be leveraged to develop and analyze interpretable dynamics that govern the population’s gameplay. By simulating the dynamics, equilibrium states of the game will be computed and translated back to the original optimization problem’s objective. The methods developed will be tested on modern AI applications, such as the nonconvex optimization problems of training and certifying neural networks. This project is inherently cross-disciplinary, drawing on tools from dynamic systems and control, computer science and programming, and advanced engineering mathematics. As such, all students with backgrounds and interests in any of these areas are encouraged to apply.

Faculty: Eltahry Elghandour, ME

Email: eelghand@calpoly.edu

Accepted Projects Modes: In-person, Hybrid, Fully Remote

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This study investigates the bulletproof capabilities of Kevlar composites fabricated using advanced manufacturing and additive manufacturing techniques. In this research, conventional ceramic armor plates will be replaced with lightweight plates made from multifunctional composite materials, with the goal of increasing impact resistance while reducing overall system weight. Continuous Fiber Reinforced 3D Printing (CFRP) is employed to manufacture Kevlar-reinforced polymer matrix composites (PMCs) with optimized structural configurations specifically designed for ballistic protection. The primary objective is to evaluate the impact resistance of these multifunctional composites against 9mm projectiles through systematic experimental testing. A series of ballistics tests will be conducted using a high-velocity gas-gun system to launch 9mm projectiles at varying velocities and angles. Testing will take place at the San Luis Obispo Range, where projectiles will be fired under controlled conditions at multiple impact angles. In addition to manufacturing parameters, the influence of different sawing processes applied to the raw Kevlar composite materials will also be investigated. Variations in cutting methods can affect fiber integrity, edge quality, microstructural damage, and residual stress, all of which may significantly influence ballistic resistance and energy absorption performance. Key parameters, including fiber volume fraction, reinforcement strategy, infill patterns, and post-processing methods, will be analyzed to determine their effects on energy absorption, penetration resistance, and failure mechanisms. The experimental results will be evaluated to assess energy dissipation, damage propagation, and overall structural integrity across different composite configurations. By replacing traditional ceramic plates with advanced multifunctional composite plates and integrating additive manufacturing techniques with rigorous ballistic testing, this research aims to contribute to the development of lightweight, high-strength armor systems. The findings offer potential advancements in protective gear and defense applications through improved performance, optimized processing methods, and enhanced impact resistance.

Faculty: Eric Espinoza-Wade, ME

Email: erwade@calpoly.edu

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Infants who sustain asymmetrical perinatal brain injury (APBI) have a 30-68% risk of developing cerebral palsy (CP). Hemiparesis (motor deficits on one side of the body) are typical of CP, and are not clinically detectable prior to six months of age. However, earlier detection could lead to early physical therapy, and improved quality of life as these infants age. The current study is designed to investigate our ability to detect asymmetry in children with APBI under three months old. We hypothesize that quantitative analyses of infant movements using computer vision and sensors placed on wrists and ankles will result in sensitive measures of asymmetry. If successful, these outcomes may shift clinical science, justifying early rehabilitation for infants in the neonatal intensive care unit (NICU). We are seeking students to refine existing code used to analyze videos of infants, to analyze wearable sensor (accelerometer) data, and to develop new measures of motor asymmetry. This project will involve collaboration and interaction with the physical and occupational therapists from Chapman University (Irvine, CA) and NICU staff from the Children’s Hospital of Orange County.

Faculty: Michael Holden, ME

Email: meholden@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Small autonomous boats are frequently used to map depth and bottom contours in bodies of water such as lakes, estuaries, and oceans. This project will use an autonomous boat (n3m0) to investigate the usefulness of commercially available depth sensors. The goals are to advance knowledge regarding optimal sensor specifications and to give the student practical experience developing and operating an autonomous vessel. The project will focus on quantifying error due to lag in a low-update rate (1 Hz) commercially available depth sensor. After measuring the bottom contour at various speeds, downsampling techniques will be used to match the depth soundings with the GPS position updates, and depth errors will be quantified. The results will be used to suggest solutions (both software and hardware) to minimize errors in the depth measurements. Finally, the solutions will be trialed and the project results will be presented. Note: this project’s field work will be at the Vallejo campus of Cal Poly Maritime and other nearby locations and is best suited to a student living in the San Francisco bay area in the summer. Both Solano and San Luis Obispo students are encouraged to apply.

Faculty: Benjamin Lutz, ME

Email: blutz@calpoly.edu

Accepted Projects Modes: Fully Remote, Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Generative artificial intelligence (AI) tools such as ChatGPT are now commonly used by engineering students to support problem solving, concept explanation, and idea generation. These activities are central to engineering work, and so how students engage with AI during engineering tasks is critical to how they develop a sense of self as an engineer. We therefore argue that student use of AI is not simply a technical or ethical question, but an identity relevant practice that shapes how students understand learning and professional development. This research project examines how engineering students decide when and how to use AI, and how those decisions relate to their developing engineering identities. In particular, we focus on engineering identity—students’ beliefs about their competence, interest, and recognition by others—to understand how AI use shapes learning experiences. The Summer Undergraduate Research Program (SURP) project will support an undergraduate researcher in helping to establish the theoretical and methodological foundation needed to launch this study. During the summer, the student researcher will engage in guided literature review, qualitative research training, development of interview protocols, and pilot data analysis. The outcomes of this project will include a refined conceptual framework, a finalized interview protocol, and a research poster presented at the SURP symposium. This work will contribute to ongoing research on AI in engineering education and provide the student with hands-on experience in education research methods and scholarly dissemination. This project provides an opportunity for students interested in engineering, education, or human-centered research to develop research skills while contributing to timely questions about learning and professional development in engineering.

Faculty: Eric Ocegueda, ME

Email: ocegueda@calpoly.edu

Accepted Projects Modes: In-person, Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Densely packed granular materials are ubiquitous across geo-mechanical (e.g., gravel), agricultural (e.g., crops), and many more fields. In each of these fields, granular materials support external loads through complex, heterogeneous networks of inter-particle forces known as force chains. The structure of these chains depends on particle scale physics and directly influences macroscopic behaviors such as stability and elasticity. The standard computational tool for measuring force chains is the discrete element method (DEM), which provides detailed particle-level data but at a high computational cost. Recently, researchers have explored data-driven alternatives to DEM, with Graph Neural Networks (GNNs) showing particular promise due to their ability to naturally represent granular assemblies as graphs, where particles are nodes and contacts are edges. The proposed SURP project aims to build off a previous SURP project, where an in-house GNN model for force chain prediction was developed, to test the GNN model and prepare for experimental comparison. Specific goals are (1) testing the GNN model’s sensitivity to various inputs, (2) exploring the GNN model’s accuracy with different error choices for training, and (3) researching the Granular Element Method (GEM), an experimental approach to predict force chains. One SURP student (from ME, AERO, CE, or other related disciplines) will work on familiarizing themselves with the developed GNN code in Python, explore machine learning literature for different loss function choices, implement sub-routines to test chosen errors and create an in-depth plan for future experimental testing of the GNN model with GEM. Students should indicate in their application if they are interested in continuing work on the project during the academic year for independent study course credit.

Email: rsrithar@calpoly.edu

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Carbon nanotubes (CNTs) are widely known for their exceptional mechanical, electrical, and thermal properties, making them attractive materials for improving the performance of traditional laminated composites. However, the effective dispersion of CNTs in polymer matrices remains a critical challenge, as poor dispersion can lead to agglomeration, reduced material performance, and weak interfacial bonding. This research study aims to investigate how CNT geometry, surface functionalization, and weight percentage influence CNT dispersion manifestations and interfacial behavior in epoxy-based nanocomposites fabricated under controlled processing conditions. In this study, epoxy panels reinforced with different types of CNTs including pristine straight, pristine helical, and functionalized straight CNTs will be fabricated using a FlackTek mixer at multiple CNT loadings (e.g., 0.025 wt% and 0.05 wt%). Advanced contact angle measurements and drop shape analysis will be employed to characterize the surface wetting behavior of the cured nanocomposite panels. Static and dynamic contact angle measurements, including advancing and receding angles, will be used to quantify wetting behavior, contact angle hysteresis, and spatial variability across panel surfaces. These wetting metrics will serve as indirect indicators of surface heterogeneity and CNT-epoxy interfacial compatibility, which are influenced by CNT dispersion quality and interaction with the polymer matrix. By systematically comparing CNT geometry, functionalization, and loading, the research will identify trends linking CNT characteristics to wetting behavior and surface uniformity. The results of this work will provide practical insights into CNT selection and processing strategies for epoxy nanocomposites and inform future optimization efforts for high-performance structural nanocomposite applications.

Email: xwu@calpoly.edu

Accepted Projects Modes: In-person

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

This project focuses on the design, analysis, and experimental validation of novel foldable drone frame mechanisms for the Poly UAS Club to enhance portability while maintaining structural integrity during flight. Existing foldable arm designs exhibit limitations in rigidity, locking reliability, and structural efficiency. To address these issues, several new distinct folding arms will be designed, developed, and analyzed using CAD modeling, finite element analysis (FEA), and kinematic simulations to evaluate stiffness, compactness, locking reliability, weight efficiency, and fatigue resistance. The highest-performing designs will be fabricated using rapid prototyping methods and subjected to load, bending, and cyclic durability testing to validate structural performance and fold-cycle endurance. The expected outcome is the identification of an optimized foldable frame design that preserves flight performance while maximizing portability and long-term durability across repeated deployment cycles.

Email: myekanif@calpoly.edu

Accepted Projects Modes: Fully Remote, Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Atomic force microscopy (AFM) and nanoindentation are powerful tools for characterizing the properties of heterogeneous materials that contain small particles. In the context of the 2026 Summer Undergraduate Research Program (SURP) project, the researchers will use ANSYS finite element analysis (FEA) to model AFM interactions. The heterogeneous material being studied consists of nodules of varying sizes and embedment depths placed within a membrane. The aim of the proposed study is to develop a model of the nodule-membrane structure in ANSYS and to simulate AFM indentation to extract mechanical properties. The ultimate goal is to create a simplified model that can assist in interpreting AFM data from heterogeneous and soft samples, such as biomaterials and polymers. Utilizing AFM and FEA to characterize modified polymers with semiconductor inclusions presents significant opportunities for the semiconductor industry, which has remained largely unexplored. Fillers have great potential to enhance the performance of rubber nanocomposites, especially within the security industry. Furthermore, studying biological samples at the nanoscale is crucial for the scientific community, as developments in AFM technology have enabled the characterization of live cells and tissues, particularly in cancer research.

Email: mzare@calpoly.edu

Accepted Projects Modes: Hybrid

Application Link (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Despite growing participation in STEM degree programs, women remain significantly underrepresented in computational and tool-intensive engineering roles, comprising less than one-third of the STEM workforce [1, 2, 3, 4]. Research shows that this gap is mainly influenced by access, learning environment, and confidence development [5, 6, 7]. Limited exposure to computational tools and mathematically intensive engineering problems can restrict persistence in relevant disciplines, reinforcing the “leaky pipeline” in STEM. Recent studies suggest that research-motivated and hybrid instructional models can help mitigate these disparities. Guo et al. [8] and Seymour et al. [9] report that structured course resources and research opportunities enhance women’s persistence, while peer interactions more strongly influence men’s STEM commitment. Hybrid education models, combining in-person and online instruction, also increase participation of historically underrepresented groups [10, 11]. Building on these findings, the PI proposes the development of an open-access, Python-based computational toolkit that combines structured instruction with research-oriented engineering problems. The toolkit compares analytical and computational solutions for various problems, with interactive features that let users adjust parameters and plot the outcomes. It also includes tutorials showing how to implement these problems on an open-source platform with multimedia support. Selected problems will be paired with laboratory demonstration experiments to connect computation with real physical systems. The platform will include interactive feedback features, allowing students to ask questions and provide input either anonymously or in person. This hybrid learning approach is designed to encourage participation among historically underrepresented groups.

Email: zhangas@calpoly.edu

Accepted Projects Modes: In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Tensegrity structures are composed of stiff rods and elastic cables suspended in a flexible tension network. In particular, the biotensegrity model proposes that all biological systems exhibit tensegrity-like characteristics across multiple scales, ranging from the cellular level to the musculoskeletal system of tendons, ligaments, and fascia, to the human body as a whole. Compared to the traditional biomechanical models used in exoskeleton design, it can be a more accurate representation of how motion emerges from natural forms, but further work is needed to fully understand the heterarchical nature of human anatomy. This project will focus on implementing feedback control on an upper-limb tensegrity exoskeleton to evaluate a user’s muscle efforts in rehabilitation. The undergraduate researcher will integrate sensors (e.g., IMUs) to track a user’s motions, adjust the exoskeleton motion in response, and compare the effects on muscle effort (i.e., EMGs) to optimize the design and control of the system.

Email: truiz01@calpoly.edu

Accepted Projects Modes: Hybrid, In-person

Application Link  (Cal Poly Maritime students should contact bbenson@calpoly.edu)

Reductions in cost and increases in efficiency of high-throughput sequencing over the last decade have substantially increased the feasibility of studies that sample and sequence environmental DNA (eDNA) to understand biodiversity. In marine science especially, eDNA now provides a major source of data on the composition of microorganisms interacting in a given environment, which form the base of the food web and have significant ecological impacts across trophic levels. Such community interactions form complex networks that are estimated from data using statistical methods originally designed for microbiome research. Studies often seek to address how communities change locally across space and time by sampling longitudinally and across environmental gradients (e.g., depth) and estimating multiple networks. In practice, statistical estimation of each network is typically performed independently as well as at multiple taxonomic levels (e.g., species, genus), but this approach ignores dependence in the data in two important ways. First, interactions are necessarily constrained across taxonomic levels: if two species of different genera interact, the corresponding genera must also interact; and if two genera interact, there must be at least one species-level interaction. Second, community structure is locally stable within microclimates and across short time intervals, so a greater degree of similarity is expected between “nearby” communities. Ignoring dependence produces inconsistencies in estimated networks that undermine downstream inferences. This project leverages prior approaches for incorporating dependence constraints into estimation of multiple networks (e.g., Guo et al. 2011, Danaher et al. 2014) to develop methods of estimating ecological associations that properly account for taxonomic hierarchical structure and local stability. The proposed approaches will be benchmarked on synthetic data and applied to two 16S rRNA amplicon datasets from California coastal ecosystems. Students on this project will contribute to software, simulation-based benchmarking, and data applications.