2026 ASEE Annual Conference & Exposition

Development of an Advanced AI Laboratory Module for Deployment on an Edge IoT Device for a Targeted Application: Microplastics Classification

Applying AI to solve real-world problems requires students to understand and use AI tools in meaningful ways during their STEM education and curriculum. Students must not only be able to use the tools but create solutions to specific problems that might require training and testing appropriate AI models suitable for the application at hand. Although students demonstrate that they are effective users of AI tools, including generative AI tools, for educational and learning purposes, developing and implementing AI solutions deployable in hardware for a specific problem would constitute a unique technical skill valuable for the STEM workforce that would require such skills from multiple technical fields.

In this paper the development of a laboratory exercise for a technical elective course on Internet of Things open to multiple engineering and engineering technology disciplines is described. The lab exercise is dedicated to training and testing a lightweight AI model that is then implemented on an edge IoT device, such as a Raspberry Pi board, for detection and classification of microplastics as the targeted application. This lab exercise is expected to significantly enhance AI learning and deployment skills of engineering, engineering technology, and computer science students, and while working on multidisciplinary teams, that can then be transferred to the STEM workforce.

In particular, this work presents a mechanism for image-based classification of different microplastics, such as beads, nurdles, fibers and other small plastic pieces, using a publicly available or student-acquired dataset. The exercise entails training a lightweight machine learning model offline that is then deployed on a Raspberry Pi microcontroller to perform real time classification. This new exercise builds on a previous exercise developed by the team which uses an already-trained AI model, such as the Google Gemini, on a Raspberry Pi board, with the main purpose of detecting and classifying known objects in a scene. The new lab exercise serves as an advanced AI module on an edge IoT device that assists students with learning how to train an AI model using a software tool, and integrating it with hardware for the specific classification problem for which it was developed. The students then validate their model with new data collected through a camera integrated onto the Raspberry Pi module.

The output of the AI model represents the determination of the type of micro plastics (detection and classification), output confidence score, and results of analytics which are all published to a Web-based dashboard. The dashboard provides educators and students with visual feedback that includes component classification, performance metrics, confidence score, number of instances detected, as well as any other output of interest. The work not only presents AI and IoT on the edge but also establishes the laboratory framework for teaching embedded machine learning, edge computing , and IoT that are suitable for upper division engineering, engineering technology, and computer science students, helping students learn the core concepts of AI, IoT, and the cloud, through hands on project-based learning as well as peer-learning , which the students can apply to their future projects or transfer to the workforce.

Authors

Dr. Mehrube Mehrubeoglu http://orcid.org/https://0000-0002-5927-8408 Texas A&M University - Corpus Christi [biography]

Dr. Mehrubeoglu received her B.S. degree in Electrical Engineering from The University of Texas at Austin. She earned an M.S. degree in Bioengineering and Ph.D. degree in Electrical Engineering from Texas A&M University. She is a professor of Electrical Engineering at Texas A&M University-Corpus Christi currently serving as the Program Coordinator for BS in Electrical Engineering and MS in Engineering programs. Her research interests are multidisciplinary in nature and entail applications of imaging modalities (hyperspectral, thermal, color) for image segmentation and object classification using classical image processing methods and machine learning/deep learning models. She is also interested in engaged student learning using engineering pedagogies, advanced technologies such as IoT and the cloud, with consideration of sustainable design and professional skills.
Muhammad Ibrar Texas A&M University - Corpus Christi [biography]

Muhammad Ibrar received his B.S. degree in Telecommunication Engineering from Balochistan University of Information Technology, Engineering & Management Sciences, Pakistan. He is currently pursuing a Master of Science in Engineering with a concentration in Electrical Engineering at Texas A&M University–Corpus Christi, where he serves as a Graduate Research Assistant. Prior to his graduate studies, he worked as a Hardware Engineer at the Government Innovation Lab in Pakistan, gaining hands-on experience in embedded systems, circuit design, automation, and IoT applications. His research interests include computer vision, embedded systems, deep learning, edge AI, and IoT-enabled intelligent sensing and monitoring systems.
Dr. Lifford McLauchlan Texas A&M University - Kingsville [biography]

Dr. Lifford McLauchlan is an Associate Professor in the Electrical Engineering and Computer Science Department at Texas A&M University - Kingsville, and has also worked for Raytheon, Microvision, AT&T Bell Labs, and as an ONR Distinguished Summer Faculty at SPAWAR San Diego, CA. He has over 85 publications covering areas such as IoT, adaptive and intelligent controls, robotics, an ocean wave energy converter, green technology, education, wireless sensor networks and image processing. He is a co-inventor on 3 US patents related to control systems. Dr. McLauchlan is a senior member of ASEE and is the current Program Chair for the Ocean and Marine Engineering Division (OMED); in prior years, he was an officer for the OMED as well, including 2012-2014 Chair, 2014-2016 Past Chair and 2023-2025 Secretary/Treasurer. He is also a senior member of IEEE, and a member of SPIE, Eta Kappa Nu, ACES and Tau Beta Pi, and has served on the IEEE Corpus Christi Section Board in various capacities such as Chair, Vice Chair, Secretary, Treasurer and Membership Development Officer. Dr. McLauchlan has received the Dean’s Distinguished Service Award twice and the Dean’s Outstanding Teaching Award once for the College of Engineering at Texas A&M University-Kingsville.
Dr. David Hicks [biography]

David Hicks is an Associate Professor in the Electrical Engineering and Computer Science Department at Texas A&M University-Kingsville. Before joining TAMUK he served as Associate Professor and Department Head at Aalborg University in Esbjerg, Denmark. He has also held positions in research labs in the U.S. as well as Europe, and spent time as a research scientist in the software industry.

Note

The full paper will be available to logged in and registered conference attendees once the conference starts on June 21, 2026, and to all visitors after the conference ends on June 24, 2026