This post is also available in:
עברית (Hebrew)
Autonomous vehicles and mobile robots rely on cameras, LiDAR and other sensors to navigate safely. These systems perform well within direct line of sight, but struggle with one persistent limitation: blind corners. Whether in a warehouse aisle, a factory floor, or a road intersection, hazards can emerge from areas hidden by walls or obstacles, leaving little time to react.
A newly developed sensing system aims to address that gap by enabling robots to “see” around corners using radio waves. Known as HoloRadar, the technology reconstructs three-dimensional scenes beyond direct view by analyzing reflected radio signals. Unlike approaches based on visible light, the system operates reliably in darkness and under varying lighting conditions.
The key insight behind the system lies in the properties of radio waves. Because their wavelengths are much longer than those of visible light, they interact differently with surfaces. Flat walls, floors and ceilings effectively act as reflective panels, bouncing radio signals in predictable ways. By transmitting radio waves and capturing their reflections, the system gathers information about objects located outside the robot’s line of sight.
According to Interesting Engineering, advanced AI algorithms process these reflections to reconstruct hidden spaces in real time. In indoor trials, a mobile robot equipped with the system successfully mapped corridors, walls and even human subjects positioned beyond corners. The system does not replace existing sensors such as LiDAR, but complements them by providing an additional layer of perception. While LiDAR detects objects directly ahead, the system extends awareness into obscured areas, giving autonomous systems more time to respond.
From a defense and homeland security perspective, non-line-of-sight perception has clear implications; autonomous ground vehicles operating in urban environments, warehouses, or security perimeters could benefit from early detection of hidden threats. In crowded or complex settings, the ability to identify movement beyond walls or around obstacles enhances situational awareness and reduces collision or ambush risks.
Future development will focus on adapting the technology to outdoor environments, including street intersections and more dynamic urban landscapes. Such scenarios introduce longer distances and additional signal variability, requiring further refinement of the algorithms.
By leveraging radio wave physics and AI processing, the system demonstrates a practical way to extend machine perception beyond conventional boundaries. As autonomous systems expand into increasingly complex spaces, technologies that address blind spots may become an essential component of safe navigation.
