SBIR-STTR Award

We propose to develop the conceptual and technical design for an autonomous vehicle that can seek, detect, localize, and retrieve wounded patients from hostile situations. We plan to concentrate on three core technical areas: 1) coverage planning and autonomous navigation; 2) patient detection and identification; and 3) vehicle design for patient retrieval. Critical navigation system technical areas that will be examined include ensuring the entire search area has been checked for patients, determining which areas of potentially rough wilderness terrain are passable and impassable, and searching and navigating without an a priori map. For the patient detection and localization task, issues we propose to investigate include attempting to identify a patient that is probably not moving, is close to the ground, and is partially occluded by natural vegetation or perhaps masked by the soldier's own camouflage. Once a wounded patient has been positively identified, the final task is to safely retrieve them. Our plan is to begin with an existing combat-ready outdoor vehicle platform and modify it to support this task. This effort will focus on the necessary additions to support grasping the patient and providing them with a sheltered, medically effective transport facility. The use of an autonomous vehicle to retrieve wounded personnel from hostile conditions not only reduces the risk of harm to those whose task it is to bring the wounded patient back, but also provides some measure of protection to the patient himself, increasing his or her own chances for survival. From a commerical perspective, we believe that there are significant spin-off technologies that could be applied sooner to existing robotics and related products. Specifically, we have identified three market areas - sensor assisted tele-operation, surveillance and monitoring, and hospital operations - where the component technology that is developed for the patient recovery task could be commercializable

The proposed robotic patient recovery system consists of two vehicles configured for marsupial operation a Robotic Extraction vehicle (REX) and a Robotic Evacuation vehicle (REV) coupled with intuitive operator interfaces for both the in the field medic and/or personnel at the remote base station. These components are linked with a flexible, JAUS-based control architecture that allows the medic to operate each vehicle at different levels of autonomy, from teleoperated to fully autonomous, depending on the his assessment of the difficulty of the situation. The small, agile Robotic Extraction vehicle would be used for short-range patient retrieval, and the larger, faster Robotic Evacuation vehicle would be for long-range patient transport from the battlefield to the medical unit, as well as for transporting the Robotic Extraction vehicle to and from the field. Both vehicles contain on-board computing and sensors to allow them to autonomously navigate in rough, outdoor terrain. The Robotic Extraction vehicle also possesses a suite of additional sensors to automatically locate wounded patients. The operator interfaces for controlling these vehicles would consist of a large, table-top version for a remote medic at the base, and a small, handheld version for the medic in the field.

Benefits:
While the military benefits of this technology are clear, substantial commerical opportunities exist for not ony the complete system (perhaps modified for other domains), but also for the constituant technologies. For example, automatic search (coverage) area path creation is directly applicable to agricultural and lawncare markets where complete cutting of crops or grass is important. Similarly, the core retrotraverse capability embdded in the REX has great utility as a simple way to augment commericially available robots that rely completely on teleoperation. Using the retrotravere function, the operator could simply drive the path once (to the hostage site for instance) and then monitor the robot as it makes subsequent traversal using the stored path.

Keywords:
Robotic Patient Recovery, Autonomous Navigation, Automatic Search Generation, Automatic Patient Detection, Marsupial Vehicles