Amazon is developing drone technology to deliver goods directly to customers. It has been predicted that drones will be ubiquitous in next few years.
However, there are some key challenges that must be addressed before drones get widely adopted in the market.
These drones generally operate at low altitudes where air traffic is busiest and most unpredictable. For example, take-off and landing of commercial aircraft, test flights, private pilot activity, hobbyists, balloons and blimps, aerial advertising, float planes, emergency responders and other drones are more likely to be present within the drone’s airspace. Therefore, a drone, operating autonomously or under the control of an operator must actively avoid interference with other objects, both moving and stationary, that are present within the drone’s airspace.
Accordingly, Amazon is developing active collision avoidance systems for its delivery drones. These drones will include multiple sensors including cameras capable of capturing multiple wavelengths of electromagnetic energy including infrared and visual, acoustic sensors and multispectral sensor for the detection of objects during the drone’s operation. The use of multispectral sensors is the key innovation here. These multispectral sensors assist in better object identification. The multispectral imaging is used to determine the composition of an object by capturing specific frequencies of reflected electromagnetic energy. The polymer of the paint or similar coating may produce a unique spectral fingerprint from which an object can be identified.
Then, the drone compares its own flight plan to the trajectories of the detected objects and updates its own flight plans to minimize or eliminate the likelihood of interaction with the detected objects.
The object avoidance system is used continuously to ensure safe travel for the drone and objects within the drone ‘s airspace throughout the drone’s operation.
Publication number: US 20160247407
Patent Title: Commercial and General Aircraft Avoidance using Multi-spectral wave detection
Publication date: 25 Aug 2016
Filing date: 12 Dec 2014
Inventors: Nathan Michael Paczan; Daniel Buchmueller;
Original Assignee: Amazon Technologies, Inc.