This post is also available in: עברית (Hebrew)
With the rise of unmanned aerial vehicles (UAVs), so rises the risk of mid-air UAV collisions, whether with other unmanned craft, or worse, with craft bearing humans. This risk has not gone unnoticed, and some of the best minds of our generation are working on collision avoidance systems to combat the problem. Now, the US Defense Advanced Research Projects Agency (DARPA) has jumped into the fray.
DARPA’s Aircrew Labour In-Cockpit Automation System (ALIAS) programme recently tested and demonstrated its own plug-and-play system “designed to enable manned and unmanned aircraft to automatically detect nearby aircraft and avoid potential mid-air collisions,” UAV Vision reports.
During the demonstration, a drone was repeatedly approached by a Cessna 172G light plane from various directions. Each time, the drone used the integrated sense-and-avoid (SAA) system to detect and track the aircraft.
At this stage, the SAA system is made of a single optical camera and passive ranging equipment. Data from inputs is used to assess the likelihood of aircraft to intersect through the flight-path of the the host craft. Collision-avoidance mechanisms determine the best course of action to avoid a collision.
DARPA hopes to create a low-cost, easy-to-install, and low-maintenance system for both drones and manned aircraft.
“This successful flight test is a step toward adding external perception to ALIAS’ toolkit for advancing in-flight automation,” Dr Dan Patt, a programme manager at the DARPA’s Tactical Technology Office. “What pilot wouldn’t want to set a box on their dashboard that would provide an additional pair of eyes to alert of potential collisions? This SAA system has the potential to enable a wide range of manned and unmanned systems to safely integrate into an increasingly populated and complex airspace.”
For the next stage, DARPA will work in shrinking the system while increasing ranging capabilities, improving aircraft detection in poor-visibility conditions, and improve collision-avoidance algorithms.