Warehouse Automation, Robotics and Autonomous Trucking

Labour is the largest controllable cost in a warehouse and the driver is the largest cost of a truck, so the oldest promise in logistics is a machine that removes them. In warehouses the promise is now real at scale: Amazon Logistics, which bought Kiva Systems for ~$775m in March 2012, said in July 2025 that it had deployed its one-millionth robot, against ~1.56m employees at the end of 2024, and introduced an AI model (DeepFleet) that it said cut robot fleet travel time by ~10% (company, reported). Symbotic, which automates Walmart's regional distribution centres, reported fiscal Q3 2026 revenue of ~$721m (+~22% y/y) and a backlog of ~$22.5bn (August 2026, reported), most of it from one customer. But automation has also produced some of the industry's costliest mistakes: SPAC-era robotics companies that collapsed, grocery fulfilment centres that never reached the volumes they were built for (Kroger said in November 2025 it would close three Ocado-equipped sites and take an impairment of ~$2.6bn; Ocado later said it would receive ~$350m in compensation, reported; see Ocado), and a patent war between Ocado and AutoStore settled in 2023. On the road, Aurora Innovation began carrying freight without a driver between Dallas and Houston on 1 May 2025 (some reports say late April), then put an observer back in the cab within three weeks at truck-maker Paccar's request; by Q3 2026 it reported ~$3m of quarterly revenue, a ~$234m net loss and a target of ~200 driverless trucks by end-February 2027 (company, reported). TuSimple, once the sector's listed leader, left the US business and renamed itself CreateAI in 2024. This section explains the automation technologies and their economics, who builds and buys them, why projects fail, how autonomous trucking works as a business, and what automation does to labour. Two companion files size an autonomous-mobile-robot fleet for peak demand and give an automation and autonomy diligence checklist.