The US Department of Defense has awarded KEF Robotics funding to develop an AI-enabled obstacle detection system for US Army tethered drones.
The company declared that it is investigating the use of infrared and electro-optical (EO) camera-based systems to detect operational hazards like tree limbs and power lines by drones attached to moving vehicles.
As the unmanned system traverses a region, it is said to employ advanced cameras and machine-learning algorithms to create a "very complex" map of the surroundings.
The software company based in Pittsburgh announced that it is currently creating scenarios and training tethered drones to avoid obstacles using the new solution by using Duality AI's Falcon simulator.
After it is created and put into place, the obstacle detection system should enable tethered drones to safely complete missions both during the day and at night.
Tethered drones are unmanned aerial vehicles that are wired with power and data to mobile military vehicles. They improve the situational awareness of warfighters by offering data and images from an aerial viewpoint.
These drones are powered by the ground constantly because they are attached to vehicles, which enables them to hover above the ground for extended periods of time and continuously monitor ground assets.
Tethered drones, however, are more susceptible to low-hanging objects like traffic lights because they operate at a lower altitude than other unmanned aerial vehicles. By incorporating a sophisticated obstacle detection system, they can detect potentially harmful obstacles from a safe distance, which would increase their mission success rate.
In order to meet the US Army's budgetary requirements, KEF Robotics claims that the solution will only use high-resolution cameras rather than more advanced data-recording technologies like LiDAR (Light Detection and Ranging). The system is a straightforward, inexpensive, passive solution because it doesn't emit any radio waves or electromagnetic frequencies.
"Avoid using that $100,000 infrared camera outside. Try using the computer that you can purchase at your neighborhood hardware or computer store, the $600 EO camera, and the $7,000 infrared camera, Kitchell advised.