TORC Robotics announced that it has received a subcontract to develop an innovative sensor fusion system for the Department of Defense. The subcontract was obtained via the Robotics Technology Consortium (RTC). The sensor fusion system will considerably enhance the high-speed obstacle detection range.
Autonomous vehicle navigating on and off road
The obstacle detection, classification and prediction system can be used to improve the navigation capabilities for unmanned ground vehicles (UGVs) utilized in mission-related settings at speeds as high as 100 KPH. The system can detect and maintain various tracking statistics for all obstacles. The company will integrate these capabilities with its main autonomy framework for upcoming availability in its AutonoNav products.
Moreover, the company will design an application programming interface for the sensor fusion software, which will facilitate integration with the autonomy framework. The company will also develop a hardware model, which can be deployed on various vehicles, including the LMTV and HMMWV.
Ground Unmanned Support Surrogate - Autonomous Vehicle
The sensor fusion system combines heterogeneous and asynchronous sensor modalities via a combined probabilistic data association method to minimize false positive or negative information, which is critical for high-speed autonomous vehicle navigation. The company will attain long-range detection and classification via a mix of radar, LIDAR and vision technologies from Smartmicro, Velodyne and Ibeo. It will then evaluate fusion and sensor performance at high-speed under a number of man-made weather conditions to incorporate dense fog, rain and snow at the Virginia Tech Transportation Institute Smart Road.
TORC Robotics will use software that was originally developed under a DARPA SBIR and currently undergoing additional research on the Collision Prediction project. TORC will use the obstacle prediction and software architecture to fulfill the project requirements. A drive-by-wire controlled ground robotics vehicle called ByWire XGV will be used for this project.