Amazon recently launched its millionth robot, which was delivered to a fulfillment center in Japan. The company is solidifying its position as the world’s largest manufacturer and operator of mobile robotics. The company also plans to launch an entirely new artificial intelligence system to make its entire fleet of robots smarter and more efficient.
A new generative model of fundamental artificial intelligence, called DeepFleet, will coordinate the movement of robots across Amazon’s fulfillment network and help reduce the fleet’s travel time by 10 percent, helping to deliver customers’ packages faster and at lower cost.
Amazon says DeepFleet should be thought of as an intelligent traffic management system for a city with congested streets filled with cars. In the same way that a traffic system helps reduce wait times and create better routes for drivers, the new AI system will coordinate the movements of robots and optimize their navigation in fulfillment centers.
Built using the company's large datasets of inventory movements across facilities and AWS tools, DeepFleet is built on artificial intelligence that learns and improves, so it will continue to find new ways to optimize robots over time, Amazon says.
Recall that earlier there were media reports that Amazon wants to start testing parcel delivery robots. To this end, the company is completing the construction of a "humanoid park", where the tests will be conducted. Before that, the company also introduced the Vulcan robot, which learned to feel touch.