Lithium-ion batteries have the potential to serve as an effective solution to the world’s ever-increasing need for energy, but with the accumulation of end-of-life batteries, proper waste management plans are lacking. This present study focuses on optimizing the hydrometallurgical recycling of lithium iron phosphate batteries using Methane Sulfonic Acid as a lixiviant. Key parameters such as pulp density, stirring speed, time, temperature, and lixiviant and reductant concentrations were altered, with the end goal of creating the most optimal set of parameters to increase leach efficiency. 100% leach efficiencies were achieved for all elements at ambient temperatures, 15 minutes, and without the use of a reductant. As such, this hydrometallurgical process has the potential to be a viable candidate to effectively separate all metals from the leaching liquor. Future studies will focus on the recovery of individual elements of lithium and iron from the metal leached solution via precipitation methods involving NaOH. The addition of NaOH will change the pH of the solution and result in the production of sodium methanesulfonate, which can then be separated into Methane Sulfonic Acid and NaOH via Bipolar Membrane Electrodialysis.

Impaired gait is a major debilitating symptom of neurological disorders such as Parkinson’s disease. However, motor symptoms fluctuate throughout the day, so clinical visits often fail to capture these short-term changes. Assessing shorter-term gait performance could enable timely interventions to mitigate symptoms, such as optimizing the frequency and dosage of medications. Existing clinical and consumer gait-tracking solutions are restricted to a single location, only capture limited data that lack sufficient fidelity, or suffer from prohibitively high costs. In addition, current designs often fail to resolve patient concerns regarding long-term comfort and the stigma associated with highly visible devices. There is a significant need to develop portable, affordable, and patient-friendly wearable sensors to monitor gait in a clinically meaningful way. GaitWay is a system that incorporates sensors on the wrists & ankles, and a mobile app for real-time, accurate gait analysis. The modular design adapts to the patient's environment, where users can wear the sensors as strap-on bracelets or sensor-integrated socks. This flexible approach significantly improves aesthetic appeal and wearability, offering patients greater convenience and discretion. GaitWay offers a more competitive price, direct consumer access, and is specifically designed for patients.

Poor adherence and lack of supervision limit the effectiveness of hand rehabilitation after injury. Therefore, this work introduces the Pinching Interface for Neuromuscular Conditioning with Haptics (PINCH). PINCH is a compact haptic device with an integrated force sensor that provides vibrotactile feedback when a target pinch force is obtained. A preliminary user study (n=5) was conducted to evaluate the device, in which participants completed 20 pinch exercises with and without haptic feedback. Force output was more consistent with vibrotactile feedback, whereas participants showed greater variability without it. These preliminary findings suggest that real-time haptic guidance may improve exercise quality. These preliminary results show that PINCH has the potential to enable objective measurements and feedback, while increasing patient compliance and participation in rehab to improve long-term functional outcomes.

We participated in the MITRE Embedded Capture-the-Flag (eCTF) competition where we designed and implemented a Hardware Security Module (HSM) that would securely store and transmit files between the devices. We programmed the TI MSPM0L228 in C to function as the HSMs that store and transmit the data. Each of the PIN-protected boards have special permissions – receive, write, read – and they are only allowed to perform actions they have the permissions for. We designed the system so that it would not perform any actions without the correct pin, any actions without proper permissions, or receive files that are maliciously generated. To defend against these scenarios, the key for encryption is derived from the permissions and PIN of the HSM through PBKDF2. To protect the data, we sign the data (using Ed25519) to prevent tampering and we encrypt it (using ChaCha20-Poly1305) so that others – without proper knowledge – cannot correctly read nor write. Without the proper PIN knowledge and permissions, the key will be invalid, preventing them from correctly performing the actions. These protections allow us to meet the security requirements of the HSM module for eCTF. This research was conducted under supervision of Professor Nader Sehatbakhsh.

Tuberculosis remains the world’s deadliest infectious disease, with increasing mono-drug resistance limiting the effectiveness of standard treatments. Although combination drug therapies can combat resistant strains, seeing which multi-drug treatments will work remains difficult. This project asks whether Drug Susceptibility Testing (Tre-DST) using the DMN-trehalose (DMN-Tre) fluorescent probe can detect multi-drug susceptibility in mycobacteria without expensive instrumentation. DMN-Tre is incorporated into the mycomembrane of metabolically active mycobacteria through the Ag85 enzyme complex, allowing fluorescence-based detection. Mycobacterium smegmatis and tuberculosis cultures were treated with first-, second-, and third-line drugs at varying concentrations to determine lethal thresholds. These limitations and previously established treatments allowed combination therapies to be examined in vitro. Cells were then labeled with DMN-Tre, and viability was measured using optical density to confirm Tre-DST and flow cytometry measurements. DMN-Tre proved to be a useful tool to determine the effectiveness of combination therapies due to its fluorescent labeling of active mycobacteria. Increased drug concentrations correlated with decreased fluorescence and optical density, indicating reduced viability. This project demonstrates the potential of Tre-DST as a rapid and cost-effective diagnostic tool for detecting TB in untreated and drug-treated conditions, improving access to timely diagnosis for underserved populations.

Metastasis is the process in which cancer cells spread from a primary tumor to form new tumors in other parts of the body and is heavily influenced by the physical process of cancer cells circulating through the circulatory system. Circulating tumor cells must undergo significant mechanical deformation to traverse capillary-sized vessels. Although recent studies suggest this mechanical deformation can enhance tumorigenicity in cancer cells, there are significant knowledge gaps in the epigenetic mechanisms driving this phenotypic shift. The epigenetic state of 4T1 breast cancer cells was examined to investigate whether mechanical deformation induces stable epigenetic programming in 4T1 cells, potentially priming the cells for a highly metastatic state. A microfluidic device was utilized to mimic capillary constriction, and 4T1 cells were subjected to controlled mechanical squeezing with optimized channel dimensions to maximize mechanical deformation. Immunofluorescence profiling of histone modifications involved utilization of the histone marks AcH3, H3K4me3, H3K9me3, and H3K27me3. The resulting findings provide insight into whether the mechanical deformation that occurs in cancer cells during passage through capillaries serves as a direct stimulus for epigenetic reprogramming, enabling the transcriptional adaptability required for metastatic colonization.

Software architecture can be one of the most difficult things to understand as a developer. With vague documentation and any possible naming conventions or organizations, understanding the inner workings of a system and how different modules work together can be a chore. Evolutionary coupling is the tendency for distinct parts of a codebase to change together frequently over time. In this study, we aim to demonstrate that evolutionary coupling crossing module boundaries is a valuable metric for understanding a system’s architecture. To validate this claim, we developed a static analysis technique that scans Git commit histories to pinpoint evolutionary coupling at the method level. We confirmed the accuracy of this technique by testing it on experimental repositories with deliberate coupling, followed by an application to three real-world C++ open-source projects: Autoware, qtbase, and MongoDB. After investigating the results, we found that evolutionary coupling correlates with architectural flaws, and may also provide weights for the trade-offs of refactoring—specifically regarding human readability, fragile legacy code, and inheritance structures.