MIT's SANDO drone navigation system avoids collisions in disaster zones using speed-limited obstacle models

MIT researchers developed a drone navigation system that creates collision-avoidance corridors through mathematical guarantees tied to obstacle speed limits, with real-world testing showing zero colli

MIT researchers developed SANDO, a drone navigation system that creates collision-avoidance safety corridors using mathematical guarantees based on obstacle speed limits. The system requires a reliable upper limit on obstacle movement speed to function and was tested across 12 real-world drone flights with zero collisions. SANDO’s lead researcher is Kota Kondo, who earned a doctorate in aeronautics and astronautics at MIT, and the system was published in IEEE Transactions on Robotics.

SANDO’s safety guarantees depend entirely on obstacles adhering to the pre-specified speed limit. If obstacles exceed this limit, the system fails to prevent collisions. The 12-flight test results do not constitute proof of readiness for all disaster scenarios, as the system’s mathematical guarantees assume obstacle speed limits remain stable during operation.

This technology could reduce human risk in drone emergency response during unstable disasters by enabling collision-free navigation without requiring real-time obstacle mapping. However, its practical impact depends on disaster zones maintaining predictable obstacle speed patterns—conditions that may not hold during rapid or chaotic events. What matters next is whether obstacle speed limits stay consistent in actual disaster environments, as this directly determines SANDO’s ability to operate safely in real-world conditions.

Source: New Atlas