UROL:ASC: AI-Supported Bionic Bivalves for Surface Water Monitoring based on Freshwater Mussel Response to Environmental Change

For over fifty years, biologists have explored using the response of bivalves to water pollution to benefit from them as sentinel organisms. Initial efforts used simple sensors to measure a limited set of behaviors. Recent developments in low power and distributed sensing, and machine learning, now enable a new generation of a Bivalve-based Living Sensor Systems (BLiSS) for multimodal, more comprehensive assessment that can help mitigate the presence of water contaminants at inlets & outlets of public water infrastructures, where they pose a risk to human and animal health. The project is expected to bring new insights into bivalve-microenvironment interactions and support modeling and data analytics that can establish baseline data for detecting environmental impacts to surface waters of public water systems. While native bivalves already play a key role in water clarity and quality as filter feeders, they are some of the fastest diminishing taxa on the planet and their conservation is a high priority. This project combines opportunities to use bivalves in bioremediation and assess their ecological impact as well as to raise awareness and support for conservation efforts. Bringing together citizen scientists and students, especially from underrepresented minority populations in STEM, these groups will interact around a range of interdisciplinary areas including embedded sensor systems, artificial intelligence, internet-of-things, data analytics, environmental conservation, bivalve biology and epidemiology.The goal of this project is to establish the fundamental physical and algorithmic building blocks of a bionic sensing system using the physiological and behavioral responses of freshwater mussels to their environment for the monitoring and surveillance of surface water resources in the inlet and outlet of public water systems. We aim to provide a) novel and multimodal on-body sensors for continuous assessment of bivalve behavior and physiology with measurements of the baseline responses to environmental conditions; b) portable, low-cost and environmentally robust embedded system platforms to be deployed in controlled laboratory conditions within the scope of this project and field in the future; c) a repository of novel data collected from single and cluster of bivalves under various environmental and pollutant exposure factors; and d) model-based and data-driven techniques to analyze the collected data in the presence and absence of ground truth validation of biosensor responses. This will initially result in an non-specific early warning system based on anomaly detection of environmental contaminants to trigger further environmental investigation and chemical testing by environmental protection agencies. The next stages of the project will focus on measured anomalies and a learning based adaptive data analytics strategy to provide a relatively more precise prediction/identification of particular exposure. These synergistic studies are enabled by a unique and actively collaborating interdisciplinary team with expertise in freshwater mussel biology, aquatic ecosystem epidemiology, bionic animal-machine interfaces, wireless embedded systems, sensors, novel materials, machine learning, and citizen science.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.