Amazon.com Inc. has filed a patent to predict and eliminate noises a delivery drone might make while flying in various conditions.
The patent describes machine learning systems aboard drones that would collect extrinsic and intrinsic data during flight.
Extrinsic data could include weather conditions, times of operation and surface conditions, such as "whether surfaces are wet, dry, covered with sand or snow or have any other texture ... or any other factors that may influence which acoustic energy is reflected, absorbed, propagated or attenuated within the given environment," the filing said.
Intrinsic data could include operational information like altitude, speed, rates of acceleration, physical attributes of the drone including its size, number of propellers and how fast they're spinning, as well as tracking information like latitude and longitude of the drones when the noises are generated or encountered.
The extrinsic and intrinsic data captured during flight can be used to train a machine learning system to associate a drone's operations, locations or conditions in such locations with certain noises.
The trained machine learning system could then be used to predict noises expected at a given location under given conditions and deploy "anti-noises" to eliminate them before they happen.
Patents and patent applications don’t guarantee an idea will become reality. But the growing number of drone-related patents shows Amazon is investing heavily in a technology that could upend the shipping industry.
The e-commerce giant made its first commercial drone delivery in December with a 13-minute flight carrying a shoebox-sized package containing popcorn and a Fire TV gadget.
Amazon (Nasdaq: AMZN) demonstrated its first drone delivery on U.S. soil in March at an invite-only conference in Palm Springs.
Hi! I am a robot. I just upvoted you! I found similar content that readers might be interested in:
http://www.bizjournals.com/seattle/news/2017/06/01/amazon-files-machine-learning-patent-to-eliminate.html