Colorado-based Black Swift Technologies is helping the U.S. Air Force improve drone maintenance planning with a recent grant.
The company will develop a machine learning software solution that can predict and improve maintenance plans to reduce drone system failures.
“System outages can be costly – time, money and equipment,” said Jack Elston, Black Swift CEO.
“Our solution uses unsupervised learning to detect anomalies. It uses algorithms that can be used to model an aircraft’s behavior in a variety of missions and flight conditions, and then look for instances that violate those models.”
Since many drone systems have no on-board monitoring or systematic maintenance, some users have to rely on guides printed in the operating instructions to determine the maintenance schedules – a limited solution at best.
While detailed maintenance protocols and schedules are standard for manned aircraft, small drones can suffer from a lack of information on subsystem status – critical components like servos are often open and unmonitored.
“By leveraging artificial intelligence and machine learning, we can create a smarter predictive maintenance plan for UAS vehicles,” added Elston. “In this way we can ensure that these UAS are operational in the sky, are safe for the local people, and remain operational at all times.”
How it works
Black Swift’s platform will use unattended machine learning algorithms to provide early warning and diagnosis of potential critical drone system failures. The system gathers important information from avionics data that the Air Force is already collecting. If the data is found to be insufficient, a number of monitoring nodes can be used to install on-board candidate platforms to supplement the data sets and implement algorithms for real-time analysis and feedback.
Black Swift News
Earlier this year, Black Swift won a NOAA contract to develop a GPS-denied navigation that enables BVLOS (Beyond Visual Line of Sight) operations for drones in GNSS-denied environments.
In 2018, the company announced the release of the Black Swift S2 UAS, which is described as a “tightly integrated small unmanned aerial vehicle (sUAS) system designed specifically for the needs of atmospheric and earth-observing scientific field campaigns.”
The S2 platform is the result of a partnership with NASA and Boulder Integrated Remote and In Situ Sensing (IRISS) at the University of Colorado (CU). It includes airframes, avionics and sensors specifically designed to measure atmospheric parameters such as temperature, pressure, humidity and wind ;; The system can lift up to 5 pounds of additional payload.
Jason is a longtime DroneLife employee with an avid interest in all things technical. He focuses on anti-drone technologies and the public safety sector. Police, fire and search and rescue.
Jason began his career as a journalist in 1996 and has since written and edited thousands of exciting news articles, blog posts, press releases, and online content.
Email to Jason
Subscribe to DroneLife here.