Companies and governments still don’t know about drones, whether they can become the next frontier in technology or another technology that has promise but cannot be secure and reliable.
Amazon has been the most notable company working on Prime drones, which can deliver small packages to customers. The problem with these drones is learning about objects and their environment, to avoid crashes.
The University of Sheffield might have a solution to the problem, using experimental quadcopters to map the environment, spot objects they have previously encountered and understand new objects they have never seen before.
This new AI functionality to the drones will give them more than enough understanding of where they are and what they need to avoid. It is a similar to the Google self-driving car, which maps the roads and has an inbuilt storage of objects, to better understand the environment.
Having this library of noticeable objects could be a way for transportation to understand the environment and move without the need for human interaction. It would mean Google and others sharing the information, allowing other companies to try it out.
In the air, drones will need a lot more environment warnings than a car. Birds, planes, trees and other objects will need to be noticed well before the drone passes by. It might take some time before the AI can work as well as Google’s car AI.
The University of Sheffield sees this object AI as being perfect for rescue missions, where drones can get to places humans cannot. If the drone understands the environment, it might be perfect for destruction zones, being able to spot humans, objects in the way and other issues.
Drones could be used in a multitude of different markets, governments, search and rescue, security, e-commerce and even social drones could be a thing in the next few years, as soon as governments start regulating them.