Rice University Develops EyeDAR: Infrastructure-Based Radar System to Enhance Autonomous Vehicle Perception
Key Takeaways
- ▸EyeDAR is an infrastructure-mounted radar system that captures reflected signals lost to scattering in traditional onboard vehicle radar
- ▸The system can resolve target directions 200+ times faster than conventional radar, enabling detection of obscured hazards like pedestrians behind objects or around corners
- ▸Infrastructure-based sensing complements the existing vehicle perception stack (cameras, LiDAR, onboard radar) to provide more comprehensive environmental awareness
Summary
Researchers at Rice University, led by Kun Woo Cho, have developed EyeDAR, an innovative off-vehicle radar system designed to significantly enhance autonomous vehicle perception and safety. Unlike traditional onboard radar systems that struggle with incomplete signal reflections and obscured objects, EyeDAR is a low-power millimeter-wave radar sensor mounted on roadside infrastructure such as traffic lights, road signs, and billboards. The system captures reflected waves that would otherwise scatter away, providing vehicles with a more complete and accurate picture of their surroundings.
The technology addresses a critical limitation in current autonomous vehicle sensing: conventional onboard radar systems often receive only a fraction of transmitted signals due to scattering, making it difficult to detect pedestrians, objects around corners, and other obscured hazards. EyeDAR solves this by serving as a complementary infrastructure-based sensing layer. Testing revealed that EyeDAR can resolve target directions more than 200 times faster than conventional radar designs, representing a substantial improvement in detection speed and accuracy. As autonomous vehicles become increasingly common in freight and delivery applications, enhanced sensing capabilities like EyeDAR are becoming essential for ensuring safe operations.
- The technology addresses growing safety demands as autonomous vehicles become more prevalent in delivery and freight applications
Editorial Opinion
EyeDAR represents a promising paradigm shift in autonomous vehicle safety—moving from purely onboard sensing to a collaborative infrastructure-based approach. The 200x improvement in directional resolution is impressive, but the real innovation lies in solving the fundamental physics limitation of radar scattering through strategic infrastructure placement. As AV adoption accelerates, this kind of complementary sensing infrastructure could prove essential for navigating complex urban environments where hidden pedestrians and obscured hazards remain a critical safety challenge.
