TANGRAM ROBOTICS, INC. — Department of Defense SBIR Phase I: AF203-CSO1

Description

Autonomous robots and vehicles rely on multiple sensors for many purposes. Some of the most critical sensors on these autonomous platforms are perception sensors, such is LiDAR sensors, IMUs (inertial measurement units) and depth sensors. Perception sensors are leveraged extensively for key autonomous tasks including navigation, data collection, systems monitoring and targeting.  Despite having been available for decades, the sensors used by these autonomous platforms remain difficult to integrate during device development, and difficult to keep performing optimally upon deployment.  Tangram’s software approach simplifies sensor integration and management to resolve these issues. An integration layer allows any number of perception sensors (both visual and inertial) to be connected to instantly stream normalized, time synchronized, structured data, while an application layer adds critical capabilities such as early failure detection and automatic calibration. This integration layer eliminates many of the most complex aspects of building sensor enabled autonomous platforms which shortens the amount of time required for deployment. The benefit of the application layer addresses critical gaps around sensor data integrity to ensure failsafe operation of autonomous platforms.  In order to ensure that autonomous robots, vehicles and systems perform mission critical operation, it is imperative for them to operate optimally. Tangram’s platform ensures the optimal performance of vision sensors and will better equip the Air Force to employ autonomous robots, vehicles and systems that enhance warfighter capability. Our software accelerates Air Force mission capability and ensures operational continuity of vision sensors and the autonomous systems that rely on them.