Amazon will spend more than €10 billion (about $11.6 billion) on next-generation warehouse robots across its European network, the company said at its Delivering the Future event in London on June 4.
The spec (what travels to a North American FC):
Next-gen Proteus takes plain-language commands and roams anywhere on site, where the earlier version stayed in the dock and ran on a programming interface
STARK, a collaborative bot that picks full totes off conveyors alongside staff, expands from a Barcelona pilot to 15 sites, both arriving in the first half of 2027
When automation pays: The robots are bound for Europe, but the decision a North American operator faces is the same one Amazon is making: when does automation actually pay back?
Per McKinsey, most warehouse-automation projects disappoint for one of two reasons. The payback period runs longer than the building's lease, or an already broken process gets automated and its problems scale with it. Automation rewards high, stable throughput that clears the payback inside a facility's horizon, and it punishes spiky demand and un-fixed processes.
Amazon can absorb a long payback at its scale; a regional operator weighing the same machines cannot, which makes throughput stability the deciding variable before the capex math even starts.
Across the sector: Our analysis of how far four major retailers are automating, drawn from their filings and earnings, shows the field splitting.
Walmart has automated 23 of its 42 regional DCs, Chewy says its automated fulfillment centers run about 18 to 20% cheaper per order, and P&G runs an unstaffed robotic night shift in Berlin. Kroger, meanwhile, booked about $2.6 billion in charges unwinding its centralized automation.
Amazon's 2027 spec is the bar a North American automation plan now gets measured against.
Dig deeper:
Meet Proteus, Amazon's autonomous warehouse robot - Amazon [Watch]
Getting warehouse automation right - McKinsey [Report]
