Combat Id – Combat Identification System

This paper will document the development of the Combat Identification (CombatID) System. The CombatID System was designed to create a platform agnostic payload that could be attached to any fielded Unmanned Ground Vehicle (UGV) to assist the Soldier in contingency basing operations. This paper will describe the approach taken to develop the system and provide a detailed description of the system, including sample results for individual modules. This paper will also provide insight on the evaluation of CombatID system’s performance. INTRODUCTION The Combat Identification (CombatID) System was designed to create a platform agnostic payload that could be attached to any fielded Unmanned Ground Vehicle (UGV) to assist the Soldier in contingency basing operations. First the robotic payload would need to meet the threshold requirement of reliable detection of Friendly Force and Foe personnel within a 60 meter radius, 180 degrees around the robot. We developed, tested and demonstrated a fully integrated hardware and software solution running on two robot systems and three additional Friendly Force entities. The proposed solution is designed to run through multiple classes of robot systems starting from Small UGV’s through large tactical or combat vehicles. Figure 1 shows the system installed on a TALON robot. The main components are: (1) the sensor head; (2) the RFFigure 1: CombatID system installed on a TALON robot. GPS unit and (3) on-board processing platform. The robot sensor system (1+2 in Figure 1) consists of a stereo camera, a MEMs IMU, a GPS receiver, an RF-ranging unit, and a dual-band mesh radio. All processing is done on-board on a small form-factor COTS computer (3) equipped with a lowpower GPU card. Each of the friendly (blue-force) soldiers is provided with an identical RF-GPS unit consisting of a lightweight GPS receiver, an RF-ranging unit and a dualband mesh radio. This paper will describe the approach taken to develop the system and provide a detailed description of the system, including sample results for individual modules. This paper will also provide insight on the evaluation of CombatID system’s performance. SYSTEM DESCRIPTION The system implements a layered approach to localization, detection, communication and classification. The navigation module on each robot fuses stereo-vision-based visual odometry with readings from the IMU and any available GPS readings to produce a robust localization solution and maintain it even in GPS-denied environments. To further improve robustness, multiple robots equipped with this system can use RF ranging to improve their relative positions. This step is accomplished with a distributed Kalman filter architecture. Once the robot's geo-location has been established, the friend-foe tracking module integrates this information, along with positions of other robots, video-based detections, and readings from sensors worn by blue-force soldiers to establish identities of the blue-force entities. The soldierworn RF-GPS units broadcast their positions and range information to the robots over an ad-hoc wireless mesh that Report Documentation Page Form Approved