Amazon has acquired Rivr, an autonomous robotics startup based in Zurich, according to reporting from The Information. The company did not disclose the acquisition price, but Rivr was valued at $110 million in an August 2024 funding round that included participation from Amazon and Jeff Bezos' Bezos Expeditions.
The real question is why Amazon needs another robotics company when it already operates over 1 million robots in its warehouses. The answer: those warehouse robots don't help with the final leg of delivery. Getting a package from a distribution centre to someone's front door remains stubbornly human-dependent. Rivr exists to change that.
Based on research from ETH Zurich's Robotic Systems Lab, Rivr designs four-legged robots with wheels that can navigate city streets, climb stairs, and handle the unpredictable terrain of suburban neighbourhoods. The company has just released its second generation of robots, which combine speed (wheels for efficient movement) with adaptability (legs for obstacles). Think of them as the urban delivery equivalent of off-road vehicles.
Here's what's clever about Rivr's approach: the robots work alongside human drivers rather than replacing them. A delivery driver can carry the robot in a van, then dispatch it to cover nearby addresses in parallel. This design sidesteps the operational complexity of fully autonomous delivery while still multiplying the work a single driver can accomplish. Amazon told its last-mile delivery partners earlier this week that it plans to test how these robots integrate into existing operations.
From Amazon's perspective, this acquisition fits a proven playbook. The company spent years automating warehouses, beginning with its 2012 purchase of Kiva Systems. That strategy worked: automation now assists in 75 per cent of Amazon's global deliveries, and the company deployed its one-millionth warehouse robot last summer. But warehouse efficiency only matters if you can deliver faster and cheaper at the final stop.
The business case is straightforward. Amazon can avoid hiring roughly 600,000 additional staff over the next few years by deploying robots instead. Each Rivr robot represents labour cost savings, reduced fuel consumption per delivery, and higher throughput per van. Rivr's AI models learn from every delivery, creating a data feedback loop that improves navigation and efficiency over time. This is what the company calls "physical AI": machines that improve by operating in the real world.
Amazon's rationale is fiscal. An internal analysis projects that automation will generate $12.6 billion in savings between 2025 and 2027. That's real money, not hype.
Yet the acquisition raises genuine questions about scalability and regulatory approval. Delivering packages with autonomous robots in urban areas requires navigating local regulations, public acceptance, and the messy reality of urban environments that don't fit neatly into a training dataset. Rivr has been working with major logistics partners, but deployment at Amazon scale is something else entirely.
There's also a broader tension at play. Amazon argues that its robots "work alongside" humans rather than replace them, citing training programmes that have upskilled over 700,000 employees since 2019. That's meaningful if true. But the company is simultaneously reducing workforce per facility to historic lows while shipping more volume per employee than ever before. The robots are augmenting human work, not supplementing headcount growth.
The acquisition reflects a company betting heavily that the next frontier of efficiency gains lies beyond the warehouse door. Whether Rivr's robots can deliver on that promise at scale remains an open question. But Amazon's track record with Kiva suggests the company knows how to turn a clever technology into relentless operational advantage.