photos of fusionsense team and project at Capstone

Spotlight on FusionSense

FusionSense is an open-source tool that allows for real-time collection of mm-wave radar data, developed by 2023 EE Capstone Seniors

From The ECE Current Newsletter (2023-24) article "Spotlight on FusionSense"

With increasing interest in object tracking technology for applications such as autonomous vehicles and airspace safety, the issues of size and scalability of the sensor units are currently blocking the widespread implementation of new technologies. Millimeter wave (mm-wave) radar is an attractive technology for these applications due to its high resolution, small antenna size, and relatively low cost-per-unit. However, radar systems can only detect positions of objects that are within line-of-sight of the units, and velocities that are parallel to the units’ antenna arrays. It can therefore be desirable to create a network of radar nodes, for example at a variety of points at a busy road intersection, that interface with one another to collectively provide all of the information needed to track all objects in the intersection.

Experimentation with mm-wave radar technology has been greatly facilitated by the release of mm-wave radar development boards, along with a software API for interfacing with the boards, by Texas Instruments (TI). However, the existing API only permits collection of raw radar data over some time period to be stored for subsequent analysis. Real-time access to raw data, which is crucial for innovations in technology development and deployment, is not available.

As part of a collaboration between the UCSB EE Capstone projects program and the research groups of Professors Jim Buckwalter and U. Madhow, a group of 5 EE seniors spent the 2022-23 academic year creating FusionSense, an open-source tool that allows for real-time collection of mm-wave radar data. FusionSense includes radar nodes with TI’s mm-wave technology that are each coupled to a software package that can both control the radar node and collect real-time data with minimal time lag. This system has opened the door to a host of applications, such as those that require real-time interfacing of multiple radar nodes. FusionSense is currently being used by researchers at UCSB for fundamental research in radar imaging targeting applications such as vehicular autonomy, as well as by a research group at the University of Alabama to develop real-time automated transcription of American Sign Language (ASL).

The ECE Current (2023-24) – "Spotlight on FusionSense" (page 10)

UCSB COE Capstone Projects - FusionSense

Learn More about the ECE 188A/B/C Capstone Senior Projects

2023 College of Engineering Capstone – EE and Multidisciplinary Projects