School of Computing & Engineering Projects Day

West Haven Stiles Elementary School in West Haven, CT, was constructed in 1930 and has been left unused for the past 25 years. This existing school is in the process of being renovated into a new apartment building. Our group designed a third-floor addition to be constructed for the renovation project. Our project goal was to identify the most efficient third-floor design, achieved by considering column layout, material selection, and load-bearing elements, all selected to ensure a safe and cost-effective design. 

Student Team Members: Taylor Byo, Robert Cooper, James Deady

Adviser(s): Priscilla Fonseca and Ari Pérez Mejía 

Clients: Sirius Structures, Thomas Nelson 

Our team developed a detailed construction plan for the design of a CSO drop shaft in New York City. In doing so we are responsible for coordinating and selecting equipment for each phase, performing cost estimations, and implementing a post-construction inspection and maintenance plan. We also identified potential risks, optimized task sequencing for efficiency, and ensured compliance with safety and regulatory standards, providing a complete, actionable construction management framework.

Student Team Members: Alex Vouras, Sean Walsh 

Adviser(s): Priscilla Fonseca and Ari Pérez Mejía

Clients: Delve Underground, Chris Dianora, Brian Lakin 

As part of a new combined sewer outfall (CSO) system in New York City, our team designed a drop shaft to reduce the wet weather effluent into nearby water bodies by nearly 90%. The project specified a 130-foot-deep drop shaft with a 50-foot inner diameter that is 3.2 miles connected to a 22-foot internal diameter tunnel. This design consisted of a post-tensioned roof structure, slurry wall support of excavation system (SOE), monolithic unreinforced concrete final lining, and bottom slab. 

Student Team Members: Michaela Capano, Kevin Clarke, Julia Duffy, Matthew Woska 

Adviser(s): Priscilla Fonseca and Ari Pérez Mejía 

Clients: Delve Underground, Chris Dianora, Brian Lakin

Artificial intelligence (AI) and cloud-based software are transforming how large companies connect with customers and drive growth. This project explores how large organizations use AI-powered Software-as-a-Service (SaaS) platforms to improve marketing, personalize customer engagement, and operate more efficiently at scale. While AI is widely discussed in business and technology, fewer studies examine how the technical design of these systems directly impacts marketing performance and long-term expansion. This research investigates how tools such as automation systems, data pipelines, and predictive analytics support smarter decision-making and sustainable growth. By connecting computer science with real-world business strategy, the project highlights how well-designed AI systems can help large companies scale responsibly and remain competitive in today’s digital economy.  

Student Lead: Carmine Andranovich

Adviser: Brian O’Neill

As AI has continued to develop in its public access, the existence of fabricated political content on social media platforms, such as deepfakes, has grown with it. Yet there are little reports of formal literacy training conducted by adult users. With expanding rates of political polarization between citizens in the United States, AI can be viewed as both a harmful weapon and a repairing tool for this deepening expanse. This research gains insight into adult users’ political alignments as well as their experiences and comfort levels with AI content on their recommendation feeds. It also investigates how AI can be utilized to counter extremist ideologies recognized through user-content interactions. Ultimately, the goal of this research is to understand how Artificial Intelligence shapes cognitive mechanisms and personal biases through its presence on social media.

Student Lead: Ella Berry

Adviser: Brian O’Neill

Modern supply chains are complex and often lack transparency, which can lead to problems such as inaccurate records, product mislabeling, and the concealment of unethical labor practices. These issues make it difficult for consumers to trust where products come from and how they are made. This project explores whether blockchain technology can help address these problems by improving transparency and record accuracy. To do this, I developed a small scale supply chain simulation that compares a traditional record keeping system with a blockchain inspired system that prevents past data from being altered. By testing scenarios involving faulty records, mislabeling, and hidden labor information, this project examines how blockchain based logging may reduce opportunities for manipulation and increase trust in supply chain information.

Student Lead: Conner Brooks-Chapman

Adviser: Brian O’Neill

This project presents how the technical architecture of modern cybercrimes enable industrial scale fraud operations across the world and more specifically in Southeast Asia. Organized groups operating from fortified compounds have forced over 100,000 trafficked workers to run online scams, these scams generate billions of dollars. Scams that are run consist of crypto laundering, AI deepfakes, social media exploitation, and traditional scam efforts. Despite international attention and enforcement efforts, operations continue to grow. The research investigates why this happens by examining technologies used and the borderless aspect by design, mixed with regional governance failures, and other failures, creating ideal conditions where traditional law enforcement cannot and will not succeed. By tracing enforcement efforts across Cambodia, Myanmar, and Laos, as well examining other materials such as frameworks, cybercrime laws, and policies. This project aims to identify the most critical vulnerabilities within the digital infrastructure that powers the mass operation of cyber fraud.

Student Lead: Javier Flores

Adviser: Brian O’Neill

Recommender systems empower platforms such as Netflix and Amazon by giving suggestions of movies and products to users. While these systems are great for convenience they are often “black boxes” giving little insight to why these recommendations are being made. This project will look into the different models of transparency – low, medium and high – to see how they affect recommendation performance. This will be done by running a simulation using a MovieLens dataset comparing how increasing transparency changes the precision and overall quality of recommendations. This study will examine whether making recommender systems more transparent to users comes at a cost of accuracy and how systems can be built with the balance of performance and better openness.

Student Lead: Omar Fofana

Adviser: Brian O’Neill

Evolutionary algorithms are computer programs, inspired by natural selection, that self-improve through repeat trial and error. Achieving balanced challenges in video games is essential for maintaining player engagement and, when applicable, supporting educational outcomes. The paper reviews the current research, examining how evolutionary algorithms can be used in video games to obtain balanced level design, computer-controlled character behavior, and broader forms of difficulty. The study includes a replication of an evolutionary algorithm from a game-theory research paper, accompanied by analysis and explanation. Research suggests that evolutionary algorithms can reliably find the values necessary for achieving optimal balance, although due to their complexity they may require a significant time input to code and test. They may adapt too slowly to sudden changes or overadapt in more stable environments.

Student Lead: Tamar Foster

Adviser: Brian O’Neill

This project aims to analyze the perspectives of game developers on the usage of automated debugging tools in the game industry. Game developers typically detect and fix bugs manually. However, this method is tedious and time consuming, so automated debugging aims to make that process more efficient. A survey was sent out to game developers that focused on getting their opinions on automated debugging tools being used in the game industry, in addition to experiences they may have had using them while developing games. A case study was also conducted to analyze the difficulty of learning how to use an automated debugging tool. Finally, the results of the survey and case study were analyzed together in order to make a recommendation of what needs to change with automated debugging tools and where future research should focus in order to allow them to become more widely adopted by game industry professionals.

Student Lead: Aleena Graveline

Adviser: Brian O’Neill

With artificial intelligence (AI) becoming a major tool for detecting lung cancer in medical imaging, ensuring both its reliability and its fairness has become a critical challenge. A staggering amount of AI models are able to perform well in controlled research environments but fail when confronted with “messy” data in real-world clinical environments or diverse patient populations. This thesis introduces the C.A.R.E. Framework, a set of professional standards to raise the importance of Accountability, Reliability, Equity, and Explainability in oncology AI models. Through extensive auditing based on these four pillars, the C.A.R.E. Framework can identify hidden biases, including performance gaps when comparing medical scanner brands or shortcuts the AI model takes to achieve a better score. The work in this thesis provides a standardized roadmap for developers and policy-makers alike to build better, ethically-based oncology tools that both clinicians and patients can trust.

Student Lead: Brooks Jackson

Adviser: Brian O’Neill

Artificial intelligence tools are becoming increasingly common in web design education, offering students new ways to generate ideas, solve technical problems, and streamline their creative workflows. This project explores how AI-assisted technologies influence undergraduate students who are learning web design, with a focus on how these tools affect creativity, skill development, and confidence in their own work. Through qualitative interviews with students currently studying web design, this research examines both the advantages of AI support, such as improved efficiency and idea generation, and potential drawbacks, including overreliance and reduced creative engagement. The findings aim to provide insight into how AI can be integrated into design education in ways that support learning without replacing foundational creative and technical skills.

Student Lead: Matthew Jones

Adviser: Brian O’Neill

Facial recognition technology is increasingly used in national defense and security settings for tasks such as surveillance, identification, and access control. While these systems are often viewed as objective, research has shown that facial recognition algorithms can perform unevenly across different demographic groups, leading to potential bias and misidentification. In defense contexts, these errors may carry serious ethical and real-world consequences.

This project examines algorithmic bias in facial recognition systems used in national defense, with a focus on public perceptions, ethical concerns, and accountability. Through an analysis of existing research and survey data, the project explores how bias is understood, tolerated, or challenged by different stakeholders. The goal is to better understand how fairness, transparency, and trust should be addressed when deploying facial recognition technologies in high-stakes environments.

Student Lead: Philip Kwiatkowski

Adviser: Brian O’Neill

Video game accessibility is not the focus of the current video games that are being produced. In order to make up for this, people will use adaptive tools, such as modified keyboards, controllers, etc., to accommodate themselves and be able to access the game. This study focuses on the use of adaptive tools within the video game community and the current climate of video game accessibility. This study also aims to gather opinions on adaptive tools from users and to bring more attention to the need for video game accessibility.

Student Lead: Ethan Lanier

Adviser: Brian O’Neill

This project develops a real-time telemetry analysis and research platform for Gran Turismo 7 to study driver performance at a granular level. The system captures high-frequency telemetry data, segments laps into curvature-defined corners, and constructs reproducible datasets for analysis. Machine learning models are applied to evaluate whether learned representations can explain performance variation beyond traditional rule-based heuristics. The research emphasizes interpretability, controlled experimentation, and comparison between deterministic metrics and supervised learning approaches in motorsport telemetry analysis.

Student Lead: Thomas Morrissey

Adviser: Brian O’Neill

Building trust is one of the biggest challenges startups face, especially when trying to gain confidence from stakeholders such as investors, customers, and business partners. Many startups struggle early on because they do not have an established reputation, which can make it harder for them to grow, attract funding, and compete in the market. Blockchain offers a potential solution by creating transparent and verifiable records of transactions and data that cannot easily be changed or manipulated. This kind of transparency can reduce uncertainty and make organizations appear more reliable.

Student Lead: Khang Nguyen

Adviser: Brian O’Neill

This thesis examines how the overproduction and lack of empathetic input in AI-generated music affect the audience's perceived value of human-composed music. It also acknowledges the existence of AI biases, both for and against AI in creative fields, and examines how these pre-existing biases shape audience perceptions of creative works. Existing research has found that many people prefer music labelled as human-created, even when they are unable to distinguish human from AI works. The risk here is that overconsumption and overproduction of these AI-generated pieces may diminish the perceived value of music made by humans and the time, empathy, and effort put into those creations.

Student Lead: MollyAnn O’Malley

Adviser: Brian O’Neill

This thesis explores the usage of generative artificial intelligence (GenAI) in game development and how its overuse can affect the human touch that makes video games so fun to play. While GenAI has been shown to help boost efficiency in idea generation and coding support, relying too heavily on those tools can harm creative vision and reduce originality of games. By examining how game developers and software developers currently use GenAI, this thesis aims to find where GenAI can be helpful and where it can be harmful. The goal of this research is to give clear recommendations for responsible and balanced use of GenAI.

Student Lead: Eugene (Trey) Paraska III

Adviser: Brian O’Neill

Artificial Intelligence (AI) is now capable of composing music that sounds insanely similar to music created by humans. But the question is, can music created by a machine make us feel the same emotions that music written by a person can? This project explores how listeners emotionally respond to AI generated music compared to traditionally composed music. People take part in an online listening study where they hear a mix of AI and human generated music excerpts and report their emotional reactions. By comparing the responses, this research examines whether emotional impact depends on who or what created the music. And the findings contribute to ongoing discussions about creativity, authenticity, and the continuously growing role of AI in emotional experiences and the artistic world.

Student Lead: Jace Pivarnik

Adviser: Brian O’Neill

This project explores the relatively untapped world of procedural video game content benchmarks. While benchmarks for both procedural game content and AI do exist, they are usually either limited in scope or don’t specifically adhere to procedural content in games. By taking aspects from other approaches to benchmarks, GenMark will be the first to properly evaluate a procedural content generator both computationally and by how effective it is at adhering to diverse, controllable, and qualitative game content.

Student Lead: Maximus Riat

Adviser: Brian O’Neill

This study examines the impact of the passage of the CHIPS and Science Act on the stock prices of a group of stocks in the semiconductor industry. Semiconductor stocks are compared to various stock indices, such as the DOW and NASDAQ, functioning as a control group. This is done by way of a difference-in-differences model, a quasi-experimental technique used in economics and data science. A short time period is examined before and after the passage of CHIPS Act. Then, to account for some overlap between the treated and control groups when using stock indices, the analysis is repeated using a random sample of non-semiconductor stocks rather than an index like the DOW. Lastly, a synthetic control approach is used to ensure the parallel trends assumption of difference-in-differences is met.

Student Lead: Ryan Slattery

Adviser: Brian O’Neill

As technology continues to grow, so does its influence on sports and athletes. While many major sports such as basketball and football receive lots of attention in regard to the usefulness of these technologies, collegiate golf is largely ignored. This project looks at how different performance technologies can help collegiate golfers improve and develop. Through semi-structured interviews with NCAA Division I coaches, this study explores how colleges currently use technologies in their programs, which tools they find most helpful, how technology shapes training decisions, and the overall opinions about the usefulness of technology in golf.

Student Lead: Natalie Spiska

Adviser: Brian O’Neill

This project explores how hospitals can collaborate to build machine learning models without sharing sensitive patient data. Using a method called federated learning, institutions train models locally and share only model updates rather than raw medical records. The project evaluates how privacy-protection techniques affect the accuracy and efficiency of heart disease prediction models. Several federated learning algorithms and privacy settings will be compared to understand how stronger privacy protections influence performance and training time. The goal is to identify practical configurations that balance privacy and accuracy. This research helps demonstrate how healthcare organizations can safely use machine learning while protecting patient confidentiality.

Student Lead: Jessica Theodore

Adviser: Brian O’Neill

This project aims to build a robot that can accurately translate spoken English to ASL (American Sign Language), making spaces more accessible for those that are hearing-impaired. With many studies suggesting that current forms of deaf to hearing communication are impersonal and non-inclusive, this robot aims to provide a more natural and efficient method of translation. Mimicking a human translator, the robot consists of a torso and subsequent limbs resting atop a mount that can be wheeled around or stay in one specific location. With the uncanny valley in mind, the robot features a facial screen that pairs ASL signs with animated expressions, which is installed on the head. The software is run on a Raspberry Pi 5, connected to two ESP32’s each with their independent power supply, allowing for powerful computation to support the local AI models running on the hardware. To translate, the robot captures audio, converts it to text, and then the text translates to ASL gloss tokens via a local lightweight AI model. These tokens are then used to access the MongoDB database to access the motion scripts which allow the robot to execute the signs. Through several waves of testing with Hartford Healthcare, signs were verified and tweaked to ensure accuracy. For signs that are not contained in our database, we have a fast and reliable finger-spelling system. 

Student Team Members: Evan Vastakis, Morgan Montz, Eric May, Shawn Acheampong 

Adviser(s): Chetan Jaiswal

Clients: Dr. Cameron LaMack, Dr. Rongyu Lin 

The ability to quickly and accurately identify an individual poses a significant challenge for many, including individuals managing memory or facial recognition impairments. To assist in 'putting name to face', we developed the I Know You Glasses: a pair of glasses that combines augmented reality and computer vision technology to provide real-time facial identification onto a Heads-Up Display (HUD). Our system was designed to split the load of two heavy convolutional neural network models across devices, ultimately solving the known 'pose variation issue' without compromising on speed or quality. We prioritized responsiveness and high-performance, allowing for the information on the HUD to remain up to date without the lag that plagues other real-time wearables.

Student Team Members: Andrew McCleary, Brady Galligan, Aislin Hayes, Jacob Levin

Adviser(s): Chetan Jaiswal

Clients: Aditya Gupta; Dr. Soumyashree Sahoo, HHC-CESI

PathFinder is an autonomous indoor drone designed to address limitations in traditional indoor security systems, such as fixed camera blind spots and the high cost of those permanent systems. The project’s focus was developing an affordable, self-contained drone capable of navigating commercial or residential indoor environments during times where humans aren’t present, such as overnight or after-hours security sweeps. Using onboard sensing, autonomous navigation, and pathfinding algorithms, PathFinder performs patrol routes and is able to monitor areas in a way that traditional systems cannot. Users receive alerts whenever the drone’s camera detects a human. As part of PathFinder’s initial setup, users scan their environment using their phone’s camera. Users then choose a drone-accessible location for its dock, which the drone will return to when it is not active. Finally, small transceivers are placed throughout the space to allow accurate positioning without the use of GPS. Once set up, users only interact with the PathFinder website. User interactions include changing hours of operation, temporarily toggling off the drone, and forcing a sweep. The portable nature of the system allows it to be packed up and moved to another space. PathFinder learns from past sweeps and passively constructs new routes as it operates. The path it takes throughout the space changes and the order in which rooms are visited is shuffled to avoid a predictable pattern that could be exploited. PathFinder shows how affordable autonomous indoor drones can be a viable alternative to conventional monitoring systems for homes and small commercial environments while providing a foundation for future work in indoor mapping and intelligent security systems.

Student Team Members: Daniel Vitashkevich, Grant Foody, Zakaria Merzougui, Fahim Kalange

Adviser(s): Chetan Jaiswal

Clients: Dr. Jonathan Blake, Dr. Jose Riofrio

Household food waste remains a significant environmental, financial, and health issue often driven by forgotten ingredients and unclear expiration labels. ScrapDish is an AI-powered mobile application, designed to help users reduce food waste through ingredient tracking and personalized recipe generation. The application allows users to log food items using OCR to read expiration dates and product details directly from food packaging, reducing the need for manual data entry. For fresh produce and items without a printed label, the application uses the YOLOv8 object detection model to recognize produce from images and automatically label them. Users may also log items using speech-to-text or manual input as an alternative. All items are stored in a digital pantry and organized by storage location, such as the fridge, freezer, or pantry. Based on these items, recipes are generated, respecting dietary preferences.  We designed ScrapDish with an emphasis on a simple visual design and automated logging to reduce user effort and encourage long-term engagement.

Student Team Members: Reeya Patel, Megan Mohr, Adelina Chocho

Adviser(s): Chetan Jaiswal

Clients: Dr. Kruti Shah, Dr. Dana White

We present our experience developing Tempo.AI in the School of Computing & Engineering at Quinnipiac University. The goal of our project is to make a AI driven DJ that mixes and is more accessible by using its artificial intelligence to support smooth transitions, clear song timing, and simple user control. During our fourth year at Quinnipiac, our team designed an interactive AI DJ application. We had numerous ideas to start with for this project, but moving forward we slowly narrowed our choices down until only one remained. To start with, our goal was to explore how different interface and audio decisions affect usability, including how quickly a new user can understand controls and produce a coherent mix. Before deciding upon this direction, we investigated a hardware-style controller interface and a playlist-only recommendation tool, but decided against both in favor of mixing AI assistance that keeps creativity with the user. After turning down those alternatives, we examined approaches suitable for building a small but complete DJ workflow. Having in mind a general idea of the experience we would like to design, we experimented with interaction flows and audio-processing strategies, ultimately deciding upon a three-model design. The first model supports AI-driven mixing by matching tempo, applying frequency filtering to reduce clashing elements, and adjusting pitch when needed so transitions remain musically coherent for listeners. The second model automatically segments songs into musically useful sections to support natural transition points, and the third model enables natural-language commands through a prompt box, including auto-queueing tracks based on metadata and similarity. Our poster will discuss our reasoning for choosing the AI DJ topic and the three-model design, and the challenges we encountered integrating real-time audio control with model outputs. The full description and implementation features of Tempo.AI, along with a planned evaluation comparing transitions to Apple’s and Spotify’s AutoMix, will be included in detail.

Student Team Members: Chris Anzilotti, Alex Powers, Sebastian Salazar, Ryan Fennelly 

Adviser(s): Chetan Jaiswal

Clients: Dr. Kruti Shah, Dr. Daniel Minogue

While universities increasingly depend on digital platforms to manage academic resources, administrative services, and student support, the expansion of these systems has led to challenges navigating information, causing user confusion and frequent reliance on support staff. This work introduces an AI-driven, conversation based, personal assistant designed to improve accessibility and efficiency. The proposed system leverages NLP, RAG, and secure integration with university infrastructure to deliver accurate, context-aware responses. Unlike traditional search bars or static help pages, the conversational interface interprets user intent and provides reliable, citation-supported answers. The system is also capable of referencing prior interactions with the current user to develop personalized responses based on the user's interests, while ensuring that all personal information is encrypted to protect students' privacy and comply with FERPA regulations. By simplifying routine inquiries and guiding users through complex digital portals, the chatbot reduces workload for administrative personnel while offering students and staff consistent, round-the-clock assistance.

Student Team Members: Samantha Woodburn, Tuana Turhan, Thomas Rua 

Adviser(s): Chetan Jaiswal

Clients: Dr. Lynn Byers

Right now, the computer science industry is a vital and abundant job market, supplying lucrative opportunities for those interested. However, an issue has become apparent to us: education for this field is lacking, and there are no good platforms for students to educate themselves. We have found that platforms that tout themselves as educational, such as Scratch or Code.org, do not improve a student’s cognitive ability or problem-solving skills. We have developed a solution to the problem called Compile Quest, and this study is to demonstrate that our platform educates students in a fun and effective manner. Compile Quest seeks to gamify and ease newer learners into coding through structured, level-based learning. Within Compile Quest, the player is presented with multiple planets that cover a vast array of coding concepts that would typically be taught to students from ages 10-13. Through the earlier planets, the player is informed about simpler coding concepts such as variable declaration and loop usage. As they proceed to the later planets and levels, however, they’re introduced to more advanced concepts for their age group. Once we roll out our finished product to our two demo schools, Saint Rita’s and Hamden Middle School, we aim to test how well students learn from our platform. A control group of students is asked to decipher a block of code before starting. Then, a group of students will clear the first two worlds of the game. Additionally, we will have a third group play Scratch to compare our results. Finally, the study group is asked to decipher a similar block of code.  Our goal isn’t to teach Python or Java but to teach concepts and reasoning. By the end of our game, the goal is for our users to understand core concepts and improve their reasoning.

Student Team Members: Ellie Macdonald, Diego Matayoshi, Lehan Dharmatilake, Mark Pfister

Adviser(s): Chetan Jaiswal

Clients: Dr. Christian Duncan, Vaishali Shah

Traumatic Brain Injury (TBI) is referred to as an injury that is caused by trauma to the head. These injuries can cause long term and short-term effects to regular behavioral functions. Based on the degree of TBI, some patients must schedule appointments with a neuropsychologist to assess cognitive and behavioral function. Although neuropsychologists can get data from these appointments, there is difficulty in assessing the patients in their day-to-day lives. The difficulty stems from the lack of consistent monitoring of TBI symptoms and their impact. With that, neuropsychologists are unable to provide the best care possible without consistent knowledge on the patient’s status. To address these concerns, our mobile application titled “ReMind” serves as an at-home therapy and monitoring application for patients to aid them in treatment. The app provides the user with several exercises, each of the exercises targeting one of the categories such as memory, attention, visuospatial, and phonological loop. The application will also include heart rate information for each session to collect further insights on their condition from the comfort of their homes.

Student Team Members: Billy Godoy, Michael Alvarado, Joshua Molin 

Adviser(s): Chetan Jaiswal

Clients: Dr. Joseph Fetta, Dr. Brian O’Neill, Dr. Joshua Haight

Interviewing skills are integral to furthering someone's career and a major hurdle for many. InterQu’s goal is to help its users improve their skills in interviewing and to get our users to be comfortable for their next real interview. InterQu leverages AI and visual/audio analysis to simulate, grade, and coach users through their interview skills. InterQu uses facial recognition and emotion detection technology through LibreFace, allowing interviews to be dynamic. MediaPipe’s Eye tracking and posture recognition can allow InterQu to detect whether a user is confident or uneasy. If a user performs well in their mock interview, harder and more thought-provoking questions will be given for the user to answer. In turn, if the user is struggling, easier questions that will aim to make the user more comfortable will be given instead. Through mock interview recordings, users will receive feedback from AI that will point out the specific skills that the user needs to improve so that they can practice their skills overtime. We are using OpenAI, which allows us to scale questions based on their responses to these questions, thus allowing users to have a comfortable experience while also improving their skills. A user’s interview transcript can be seen to the AI model to evaluate their response. This means, after interviews are completed, users will receive actionable feedback that they can use to help improve their skills. With our backend Supabase database, each user will have their own profile where they can save their past interviews. By showing users feedback on their responses to questions, posture, eye movements, and emotions, users can then see these specific areas of where they can improve and incorporate strategies to fix their skills. Through repetitive use and practicing the feedback provided to them, users will hone their skills and see themselves grow.

Student Team Members: Aidan Armelino, Lucas Jenkins, Hayden Lacy, Connor Ryan 

Adviser(s): Chetan Jaiswal

Clients: John Bau, Harsh Gandhi

Partnering with Lockheed Martin (Sikorsky), this project focuses on the research and development of an automated process, based on a current manual process. Our objective is to determine the impact of implementing automation technology within the company’s internal logistics and material handling operations and to provide recommendations for process improvement. By incorporating virtual simulation, robotic, conveyor, and sensor technology and programming, manual and automated time studies, process flow mapping, and comprehensive research, we have compiled a detailed feasibility report for the company.

Student Team Members: Matt Alduino, Briana Craig, Gavin Grim, Jack Krut, Julie Lawson, Melanie Martinelli, Erin Swearingen, Eli Toomer

Adviser(s): Justin Kile

Client: Lockheed Martin (Sikorsky)   

Based on the requirements of the ASME Waste Collector Challenge, the Department of Mechanical Engineering at Quinnipiac University seeks the design and development of a Quinnipiac-adapted, remotely operated, single-vehicle waste collection and segregation system. 

The system must be capable of navigating varied campus terrain and accessing multiple waste storage locations while efficiently collecting, transporting, and sorting materials into designated compartments. 

The device must operate safely without causing damage to property or infrastructure, comply with applicable operational and safety requirements, and maintain reliable remote-control functionality within the expected operational range.

Student Team Members: Tahira Jilu, Michael Gervais, Henry Wengel, Cynthia Skenderi, Erika Pinto 

Adviser(s): Cameron Lamack

Technician(s): Dennis Hanlon

Client: Quinnipiac University School of Computing & Engineering    

Tasked with the complete design, analysis, and assembly of a custom 4WD chain-driven drivetrain system, this project focuses on developing an integrated gearbox and differential solution for the Quinnipiac University Baja SAE team vehicle. The system will be engineered specifically for implementation in the team’s upcoming Baja SAE car, with careful consideration given to performance, durability, manufacturability, and serviceability under harsh off-road racing conditions.

Student Team Members: John Lindewirth, Chris George, Jorge Angulo, Nick Spence, and Carter Stevens 

Adviser(s): Cameron Lamack

Technician(s): Dennis Hanlon

Clients: Baja SAE, Jose Riofrio 

Kinetic Communications is our senior capstone project focused on designing and developing a functional, motion-based communication system capable of producing American Sign Language (ASL). In collaboration with senior students from the School of Computing, the robot translates spoken input into precise movements that form accurate ASL signs in real time. Using a combination of servo motors, stepper motors, microcontrollers, and custom circuitry, the system replicates the motion of human shoulders, arms, and hands along with facial expressions displayed on the head to execute the complex movements required for clear and understandable signing.  

Our team is responsible for the complete engineering process, including the design and manufacture of major structural components such as the head, torso, shoulders, and upper and lower arms, as well as the development of custom motor control circuits and embedded systems. Throughout development, we emphasize reliability, repeatability, and accessibility. This project aims to improve accessibility and help bridge communication gaps for members of the Deaf and hard-of-hearing community.

Student Team Members: Kieran Boyle, Nicholas Farrell, Zachary O’Connell, Alexander Michel, Cade Wetter

Adviser(s): Cameron Lamack

Technician(s): Dennis Hanlon

Client: Dr. Chetan Jaiswal

Redesign a vehicle tray for ParkPlus’s AGV systems that address the need for improved cost and performance. The new tray significantly reduces overall weight and lowers manufacturing costs while simultaneously increasing shipping efficiency, shortening production lead times, and enhancing system reliability. Additionally, to achieve these objectives, the design incorporates an innovative process and uses fabrication methods that facilitate production in a timely and cost-effective manner. The prototype that our team, The Curbside Crew, produced has primarily serves the purpose of a proof of assembly and will be made on a one-fifth scale. The completed design solution must meet all structural and safety requirements, remain fully compatible with existing AGV operations, and be finalized and delivered to ParkPlus by May 8, 2026. Our primary contact is Jake Fitzpatrick, an engineer at ParkPlus.

Student Team Members: Kevin Murace, Charlyse LaMantia, Wilhelmina Kalish, Gabe Sousa, Alex Passaretti  

Adviser(s): John Reap  

Technician(s): Chad Hanebrink  

Client: ParkPlus (Primary Contact: Jake Fitzpatrick)

The problem addressed by the RC Garbage and Recycling Collector is the need for an efficient, compact, and maneuverable system capable of collecting and sorting waste within strict spatial and operational constraints compared to the ASME Student Design Competition requirements. The system must fit within a 50 cm × 50 cm × 50 cm bounding box while navigating narrow lanes and completing multiple waste-handling functions. Key challenges include minimizing total weight, maximizing energy efficiency, and designing a reliable lifting and sorting mechanism with sufficient range of motion. The truck must integrate a durable drivetrain, a controlled robotic arm or screw-drive lifting system, and a stable chassis that prevents tipping under load. The design must balance mechanical performance, manufacturability, cost, and competition scoring criteria.

Student Team Members: Emma Petersen, Andrew Sanz, Jack Cocca, Shayne Hasipi, Donato Cuzzolino 

Adviser(s): John Reap  

Technician(s): Dennis Hanlon 

Client: John Reap

The Seal Team 6.0 project is an automated gasket dispensing system designed for PTA Plastics in Oxford, CT. The device dispenses foam gaskets individually, removes the backing liner, and presents each gasket to an existing robotic automation cell. The system is built to handle two large and four small gasket sizes while maintaining high speed, accuracy, and gentle handling. Key constraints include reliable liner removal, OSHA and ISO safety compliance, compact bench top footprint, and durability for industrial use. The design incorporates dual conveyor belts, adjustable peel mechanisms, actuator driven loading systems, and Arduino based control with safety features including a kill switch. Modular subsystems allow iterative testing and refinement before full integration. The final design transitions from PLA prototypes to carbon fiber components for strength and long-term reliability.

Student Team Members: Andrew Miller, Oliver Hanson, Nicholas Ferreira, Zachary Kayton 

Adviser(s): Cameron Lamack  

Technician(s): Chad Hanebrink, Dennis Hanlon 

Client: PTA Plastics (Oxford, CT)

The objective is to design a support equipment mechanical system capable of loading and unloading cargo to and from a receiving vehicle.  The design must hand off weight from the support equipment to the vehicle while utilizing minimal attachment points.  The design minimizes the number of personnel needed to use the equipment, must be of minimum weight for transportation, and maintains clearance to the vehicle, the cargo, the personnel, and to the ground throughout its range of motion.

Student Team Members: Andre Boily, Colin Zemlanicky, Cynthia Staroscik, Grace Foley, Owen Korper

Adviser(s): John Reap 

Technician(s): Dennis Hanlon, Chad Hanebrink

Client: Sikorsky, a Lockheed Martin Company

Team’s project consists of the construction of a handheld stone launcher that achieves a consistent 5 skips along a body of water. On top of this, the launcher will be repeatable, launch the stone at a proper 20-degree angle, and achieve the proper velocity needed to complete the distance needed to achieve 5 skips.  The stones will be made by the team with natural clay.  The launcher will use a 270-degree torsion spring along with multiple parts such as an arm and shroud made from solid plastic and other materials.  The arm will be connected to the torsion spring, will latch back to another hook connected to a trigger. The user will then pull the trigger to launch the stone through the shroud.

Student Team Members:  Sean O’Hagan, Jonathan McNulty, Andres Argueta

Adviser(s): John Reap 

Technician(s): Dennis Hanlon

Client: Jose Riofrio

The project is to design a means of testing the performance of different turbine blade configurations. The device needs to measure and record the power generated by a turbine during a measured release of water into the system. Replacement of the turbine must be simple and quick, as the device will be used by students in a QU SCE Lab setting. The frame is created from steel, and houses two water tanks, with PVC piping and valve allowing the water to flow down to the turbine and then is cycled back up to the top tank using pumps, producing values that can be used to determine which blade configuration is the most efficient.

Student Team Members: Rachel St. Germain, Matthew Sheehy, Marcus Peguero, Kyle DeRienzo, Riley Soybel

Adviser(s): John Reap

Technician(s): Chad Hanebrink

Client: John Reap

This project tasks our group with designing, building, and testing a remotely controlled waste and recycling collection device. The device must be compact, safe, reliable, and efficient while collecting and distinguishing between waste materials. The device should also be capable of real-time waste identification and sorting, be compliant with traffic laws and safety rules, operate within size/power restrictions, and serve as both a competition entry and a model for real-world urban waste challenges. The final design must fit within 0.125 cubic meter box with none of the dimensions exceeding 50cm across, deep, or high, be powered by safely housed dry-cell rechargeable batteries and be capable of operating under the competition's guidelines regarding navigation, sorting, safety, and efficiency.

Student Team Members: Shane Carter, Matthew Wordell, Ryan Santos, Nick Lorenzo, Simon Diaz

Adviser(s): Cameron Lamack

Technician(s): Chad Hanebrink, Dennis Hanlon

Client: Quinnipiac University Department of Mechanical and Industrial Engineering

The goal of this project is to develop a software system to aid students of Quinnipiac School of Computing & Engineering (SCE) in writing high qual resume such that it closes the gap between academic coursework and industry expectations. The application allows students to easily generate personalized, highly descriptive resume bullet points based on the specific courses they have completed. Simultaneously, the application provides SCE faculty with an intuitive system to map technical competencies and skills directly from their syllabi.

Student Team Members: Tyler Rinko, Paul Zegarek, Ethan Kulawiak

Adviser(s): Ruby ElKharboutly 

Client: John Bau, Assistant Dean for Student Experience and Career Development

The goal of this project is to develop an automated inventory tracking and management system for Quinnipiac University Maker Space. Currently, the facility relies on manual methods to track supplies, which can be time-consuming and prone to error. This application utilizes a vision-based object recognition model to automatically detect and record inventory levels. The system includes a web application dashboard, where users can view inventory charts and usage data. By modernizing how materials are monitored, the application is designed to enhance the experience for students and faculty utilizing the Maker Space.

Student Team Members: Jean LaFrance, Oscar Lin, Calvin Pancavage, Brandon McCrave

Adviser(s): Ruby ElKharboutly

Clients: Lynn Byers, Professor of Mechanical Engineering; Quinnipiac University Maker Space