Document Type
Masters Project
Abstract
This project developed a custom trained artificial intelligence (AI) video‑analysis agent minimum viable product (MVP) for the University of Alaska Anchorage (UAA) Applied Environmental Research Center (AERC) fisheries projects that utilize a VAKI Riverwatcher CS system. The project involved collaboration with the UAA AERC and the UAA College of Engineering (CoE) Project Management (PM) program and the Computer Science and Computer Engineering (CSCE) program. The project served as the first phase of a broader program that will expand into an all-in-one commercial product, eventually replacing the Icelandic-made VAKI Riverwatcher CS with a fully automated, American-made fish monitoring system to comply with the federal Build America Buy America Act. Using the Hugging Face Transformers as the AI library in a Google Colaboratory (Google Colab) development framework, this project proved the concept of automating the fish identification process using an object detection AI model. The project team developed an MVP utilizing AI and machine learning to identify the salmon species returning through the VAKI Riverwatcher CS to spawn up the Otter Lake Drainage on Joint Base Elmendorf-Richardson (JBER), Alaska. The AI tool achieved a Technology Readiness Level (TRL) of three out of nine. In addition to the AI tool, the project team created a development policy that the University of Alaska’s (UA) Office of Information and Technology (OIT) Governance and Risk Compliance (GRC) Office used toward formulating their own development policy for the UA network. Finally, lessons learned were collected throughout the project so other project teams can learn from this project’s success.
Publication Date
5-1-2026
Recommended Citation
Blake, Alec, "Integrating Automated Artificial Intelligence (AI) Into University of Alaska Anchorage's Applied Environmental Research Center" (2026). Student Projects for Graduate Degrees. 273.
https://scholarworks.alaska.edu/uaa_grad_stuprojects/273