Amazon has just rolled out its new DeepFleet AI system—a smart tool designed to make its one million warehouse robots even more efficient. Using generative AI, DeepFleet learns from past robot movements and adapts routes in real time, acting much like a custom traffic control centre. Operating in over 300 facilities worldwide, it moves beyond fixed paths to plot flexible, optimised routes that help avoid delays while cutting costs and energy use. With the promise of roughly a 10% reduction in travel time, this system brings a practical solution for faster package delivery, something we can all appreciate if we’ve ever waited too long for our orders.
Amazon’s journey with robotics started in 2012 and has since grown to feature specialised units—Hercules for heavy lifting, Pegasus for parcel handling, and Proteus for autonomous navigation. While the introduction of advanced robotics has boosted technical roles (for instance, a 30% increase in maintenance staff at a Louisiana facility), the overall number of workers has declined. Recent reports suggest that staffing has dropped from about 1,000 per site in 2020 to 670 today, hinting that robots might soon outnumber human counterparts. Notably, a logistics centre in Japan recently celebrated the deployment of its millionth robot, and since 2019, over 700,000 employees have received training to keep pace with these new technologies.
