Autonomous Technologies that Indicate Hostile Intent Required

This post is also available in: עברית (Hebrew)

The US Defense Advanced Research Projects Agency (DARPA) is looking for new ways to use drones and sensors to analyze human behaviors that indicate hostile intent, to help warfighters distinguish between threats and noncombatants. Current technology cannot effectively find snipers or hidden combatants, leaving the military to rely on dismounted warfighters to patrol urban areas. To limit the danger, DAPRA wants autonomous systems that detect and positively identify hostile agents before any troops come in contact with them.

The Urban Reconnaissance through Supervised Autonomy (URSA) program would use drones, integrated sensors and advanced algorithms to distinguish between threats and noncombatants. However, URSA would also measure responses to natural or created stimuli and analyze those behaviors to help deduce the intent of persons of interest.

The agency maintained URSA will require significant advances in active sensing, behavior understanding and autonomous decision-making to determine intent. Therefore, DARPA wants proposals that focus on “new algorithms and techniques to rapidly discriminate between threats and noncombatants, as opposed to sensor, effector and platform development.”

Among other key features are: Accumulating and integrating evidence over time and from multiple sources, fast processing, accurate tracking to supply actionable intelligence, and using existing hardware, software, simulation infrastructure and physical interfaces to limit costs.