Fuse2: Topic 1: Efficient Edge Inference and Heterogeneous Integration in Systems for Health and Chemical Sensing

Artificial Intelligence (AI) has led to groundbreaking progress in tasks such as image recognition, classification, speech, and natural language processing. However, the implementation of machine learning AI models is costly in terms of energy, storage, and computation, making them unsuitable for integration into resource-limited sensors. Weightless Neural Networks (WNNs) represent a distinct class of neural models inspired by the processing of input signals by biological neuron dendritic trees.

They are small, fast, and energy-efficient. This project focuses on integrating WNN-based intelligence with cardiac and chemical sensors at the point of sensing. It leverages expertise in machine learning, circuit design, and sensors to develop integrated systems for health and chemical sensing, combining the investigators’ prior work on tiny machine learning networks, ultra-thin wearable health patches, flexible circuit manufacturing, molecular chemistry, molecular biology, electromagnetics, and micro and nanofabrication technology.

Of particular interest are intelligent systems for cardiac health sensing and innovative chemistry applications.

The integration of intelligence and sensing developed in this project is expected to benefit the common public via health monitoring advances. Integration of intelligence within ultra-thin, lightweight and multifunctional wearable patches which can conform to soft and curvilinear skin surfaces is important for cardiac and other health monitoring applications.

Such health monitoring can lead to preventive health measures and personalized healthcare. Inexpensive solutions in this domain can make the use of such sensors pervasive, enhancing health equity for the masses. The chemical sensing platform developed under this project will serve as a tool to enable the promise of basic scientific discovery in chemistry and molecular biology.

The project is also expected to train a large workforce in semiconductor technologies. The joint activity between the University of Texas at Austin and the University of Texas at San Antonio involves communities underrepresented in STEM, including women and minorities, as well as first-generation college students.

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.