Perception-driven Casualty Manipulation to Enable Near-Autonomous Robotic Extraction and Evacuation

Historically, the combat medic has had one of the most important yet life-threatening roles on the battlefield. They must find, assess, treat and extract fellow injured soldiers in extremely hazardous conditions, sometimes with little or no additional support. Because of this, medics and other first responders are often injured or even killed while attending to wounded soldiers. To properly address the problem of Robotic Combat Casualty Extraction, the team of SkEyes Unlimited Corporation and Carnegie Mellon University believes that a fundamentally new, patient-centric approach needs to be pursued. Namely, that robust sensor/perception-driven casualty assessment and manipulation is the key to successful development and deployment of a complete Robotic Combat Casualty Care system. For this STTR, we propose to use perception sensors and algorithms to accurately model casualty pose and status, and to couple those technologies with robust, yet compliant manipulators, to enable near-autonomous manipulation and handling of wounded soldiers throughout the spectrum of treatment–-from field assessment of injuries to final evacuation to primary care facilities. This extraction system will be Joint Architecture for Unmanned Systems (JAUS) compliant and will be interoperable with current and future DoD unmanned platforms.