Army Applications Lab recently awarded Larimer-based KEF Robotics, an aerial autonomy developer, a 12-month, $1.5 million, phase two Small Business Innovation Research contract.
KEF will use funding to develop advanced computer vision algorithms to fly tethered autonomous drones in urban environments at high speeds.
“Our platform-agnostic, attachable approach to aerial autonomy positions us well to provide safe tethered flight in dynamic and complex environments,” Eric Amoroso, KEF Robotics cofounder principal and investigator, said. “The tethered system allows us to tightly couple the visual sensing on the ground vehicle and Small Unmanned Aerial Systems (sUAS), from which our system uses both machine learning and stereo vision techniques to probabilistically map difficult hazards like power lines and tree branches.”
Tethered drones are connected to the ground via a power and data cable and receive power from the ground to stay in the air without recharging.
The company will use machine learning algorithms, infrared and electro-optical sensors, and powerful computation to build a dense and long-range map of a drone’s surroundings.
This generates a map in real time. The map, along with KEF’s planning and navigation modules, will allow drones to autonomously avoid obstacles while moving in an urban environment.