SDIP Research Experiences for Undergraduates (REU)

Dr. Masudul Imtiaz’s research lab focuses on developing advanced ultra-low-power wearable sensor systems for real-time, non-invasive health monitoring. These sensors aim to provide accurate, continuous tracking of the wearer’s health status. Utilizing prior designs, including multi-sensor platforms, this project addresses the need for compact, personalized wearables that combine high performance with user comfort. The project’s objective is to enable edge comoputing capabilities directly within the sensor, optimizing performance within a constrained power budget. REU students will assist in designing AI prototypes for health sensing, collaborating with the Center for Advanced PCB Design and Manufacture (CAPDM) facility and AI Vision Lab to validate new sensor systems and assess their commercialization potential. The project includes creating a custom PCB for testing using ultra-low-power ARM processors with BLE communication. Students will evaluate sensor performance, durability, data quality, and reliability through human studies, supporting long-term health monitoring for proactive healthcare and enhancing the work of healthcare providers and educators.

Led by Dr. Stephanie Schuckers, director of NSF Center for Identification Technology Research (CITeR), this project develops advanced biometric sensors to enhance national security. Working alongside industry partners like the FBI, DHS, and DoD, this research focuses on next-generation biometric systems with increased robustness against spoofing and privacy vulnerabilities. REU students will participate in building flexible biometric sensors from COTS components and updating firmware to improve biometric authentication. Students will balance trade-offs among accuracy, computational cost, and memory footprint in applications such as wearable mID and fingerprint-enabled IoT devices. The project will also involve developing embedded processors on medical wearables, enhancing biometric recognition, and evaluating performance across diverse demographics.

Dr. Xiaocun Lu’s team explores biomechanics imaging for biomedical applications, focusing on the design and sensitivity optimization of near-infrared mechanical sensors. Students will work on the molecular design and synthesis of biomechanical sensors, enabling deeper detection within biological systems. REU participants will use computer simulations to evaluate molecular sensitivity, followed by synthesis and characterization of the sensors. They will optimize sensitivity and biocompatibility, enhancing sensor penetration in aqueous conditions. Students will thus gain experience in polymer synthesis, sensor design, and simulation techniques, contributing to the development of sensors tailored for biological applications.

Dr. Silvana Andreescu’s research group develops portable sensors for detecting environmental contaminants, including per and poly-fluoroalkyl substances (PFAS), funded by NSF, USDA, and the US Army. The project involves designing and fabricating sensors on PCBs with polymer recognition properties for PFAS and other contaminants. REU students will modify electrochemical sensors and calibrate them for specific contaminant detection. Students will perform data analysis and QC/QA to validate the sensors and apply them to detect contaminants in water samples from nearby Superfund sites and other sources, gaining expertise in molecular recognition, sensor fabrication, and environmental testing.

Indoor air quality, specifically the presence of airborne particles, significantly impacts human health. Led by Dr. Suresh Dhaniyala, this project develops sensors capable of detecting particles as small as 10 nm and assessing biological particle diversity in the air. REU students will explore sensor performance, optimize data representation, and develop new sensor modalities. Working within Clarkson’s Center for Air and Aquatic Resources Engineering and Sciences (CAARES), students will gain practical experience in environmental pollution management, leveraging advanced sensor technologies to monitor and analyze airborne contaminants.

Dr. Kevin Fite’s project focuses on the integration of multi-sensory technology into powered prosthetic limbs, enabling advanced control strategies for amputees. REU students will engage in sensory fusion research, combining measurements from wearable sensors with the prosthetic limb’s state data to facilitate user-driven control. This work supports the development of field-deployable prosthetic limbs, bridging the performance gap between prosthetic devices and natural limbs. Students will contribute to electronic architecture design, system integration, and the evaluation of the interaction mechanics, gaining practical experience in intelligent prosthesis control.

Under the guidance of Dr. Thomas Holsen, this project evaluates PFAS sensors for detecting contaminants in groundwater from DoD sites. REU students will compare PFAS sensor results with EPA Method 1633 standards, assessing contaminant impact on sensor response. Additionally, students will explore real-time control of a plasma reactor for PFAS removal, using sensor data to optimize reactor efficiency. This project also introduces students to the Great Lakes Fish Monitoring and Surveillance Program, a US EPA-funded initiative. Students will gain experience in environmental sampling, sensor calibration, and real-time data application for enhanced water treatment.

Dr. Abul Baki’s team is developing an AI-vision-based sensor to detect microplastics in aquatic environments, addressing the limitations of traditional monitoring methods. USB cameras interfaced with the computer and Deep Learning based object detection model track microplastic particles in a lab setup, while a Deep-SORT model determines their velocities. The next phase will enable real-time detection using a portable NVIDIA Jetson AGX processor. REU students will explore optimal processor and camera models, collaborating with Clarkson’s Center for Advanced PCB Design and Manufacturing (CAPDM) and AI Vision Lab to refine and test the system across various water bodies, developing firmware in a Linux environment for broader field applications.

Microbial populations play a crucial role in river ecosystems, and imbalances, such as those caused by cyanobacterial blooms, can pose serious risks to plant and animal life. Analyzing microbial populations using currently available techniques such as sequencing or quantitative PCR is often costly and labor-intensive, limiting their use in routine monitoring. Dr. Shantanu Sur’s group aims to address this challenge by building a prediction model of microbial abundance from the physicochemical properties of river water. These properties can be measured through existing sensor-based methods and, thus, would enable high-frequency, low-cost monitoring of river microbes. REU students will perform DNA extraction, sequencing, and bioinformatics analysis to assess the microbiome in the tributaries of the St. Lawrence River. They will also engage in modeling and data analysis to identify key predictors of microbial composition from sensor-derived measurements.

A compressed pharmaceutical oral solid dosage (OSD) form consists of a tightly bound network of particles, with its quality attributes such as disintegration, drug release, and hardness affected by its micro-scale properties. Ultrasonic evaluation offers a rapid, cost-effective, and non-destructive way to assess these properties, but extracting the micro-properties from the ultrasonic data is a complex mathematical challenge. In the REU project, a new machine learning approach, using Multi-Output Regression models and Neural Networks, will be introduced to extract these micro-properties directly from ultrasonic waveforms. Virtual tablet waveforms will be created to train and test these ML models. These models will then be used on real OSD tablets, successfully determining their micro-properties and demonstrating the method's potential for practical application.

This REU project focuses on leveraging FPGA technology for real-time AI-powered video processing and computer vision applications, particularly targeting scenarios where low latency and efficient power usage are critical. The project aims to develop optimized AI models that can be deployed on FPGAs for tasks such as real-time object detection, deepfake detection, and activity recognition. Unlike traditional CPU or GPU systems, FPGAs offer the advantage of customizable hardware that can be fine-tuned for specific AI workloads, providing high performance within a limited power budget.REU students will assist in developing and optimizing AI models on an FPGA platform using the Xilinx Kria KV260 Vision AI Starter Kit. They will explore techniques for accelerating convolutional neural networks (CNNs) and transformer models on FPGAs to achieve real-time processing speeds. The project also includes designing hardware modules, testing on real-world video datasets, and implementing efficient data transfer protocols between sensors and the FPGA board. By the end of the project, students with Dr. Masudul Imtiaz will have developed a comprehensive understanding of AI hardware acceleration and gained hands-on experience in deploying vision-based AI systems on FPGAs, making them well-equipped for careers in embedded AI and hardware design.

As VR technologies continue to evolve, the integration of real-world data into VR environments is becoming increasingly crucial for creating immersive and interactive experiences. This REU project focuses on developing APIs that allow Raspberry Pi 5 or BLE-enabled sensors to interface with the Unity VR SDK. By enabling real-time data transmission from physical sensors to a VR environment, the system will provide enhanced interactivity and realism within VR applications. These APIs will be designed to support a variety of sensors, such as motion detectors, temperature sensors, and biometric monitors, facilitating their use in diverse VR scenarios, from training simulations to interactive gaming. This will offer developers the tools to create richer and more responsive VR experiences. Multiple faculty from Clarkson, collaborating with Dr. Masudul Imtiaz, will supervise this research.

This REU project, supervised by Dr. Masudul Imtiaz, will focus on creating ultra-low-power wearable sensors for non-invasive monitoring of infant health. The goal is to develop flexible, comfortable sensors that can continuously track vital signs like heart rate, breathing patterns, and movement. REU students will work on designing sensor prototypes, integrating microcontrollers, and developing algorithms for real-time data processing. This project includes testing the sensors on a custom PCB, optimizing for accuracy and energy efficiency. Students will gain skills in sensor design, embedded systems, and data analysis, contributing to advancements in infant health monitoring technologies.