Amazon made two robotics acquisitions in five days. First RIVR on March 19—a Swiss startup making four-legged robots that climb stairs to deliver packages. Then Fauna Robotics on March 24—a New York company building kid-sized humanoid robots.
The combined moves signal something bigger than incremental logistics improvements. Amazon is betting that robots will handle the last steps of delivery—the part that’s hardest and most expensive.
RIVR: The Stair-Climbing Delivery Dog
RIVR’s flagship robot is called Rivr Two—a four-legged machine on wheels that can carry over 60 pounds of packages, travel up to 8.7 miles per hour, and climb stairs. That last capability matters most.
Amazon’s earlier Scout delivery robot program, which used wheeled sidewalk bots, shut down in 2022 because it couldn’t handle curbs, steps, or uneven terrain. Scout worked fine on flat sidewalks but failed when it reached an actual doorstep.
RIVR’s quadruped design solves this. The robot handles curbs, stairs, porches, and the real-world obstacles that killed Scout. Amazon says it will begin doorstep delivery testing with RIVR’s robots later in 2026.
The acquisition price wasn’t disclosed. RIVR, an ETH Zurich spinout, had raised roughly $25 million total, including a $22.2 million seed round co-led by Bezos Expeditions and the Amazon Industrial Innovation Fund in 2024. The company was last valued at $100 million.
Importantly, Amazon isn’t replacing delivery drivers. RIVR robots will work alongside human drivers, helping them carry packages from vehicles to doorsteps. This hybrid approach—robots assisting humans rather than replacing them—differs from the fully autonomous delivery vision that has mostly failed to materialize.
Fauna: The Friendly Humanoid
Five days later, Amazon bought Fauna Robotics.
Fauna builds Sprout—a $50,000 bipedal humanoid robot standing 3 feet 6 inches tall and weighing 50 pounds. Unlike industrial robots designed for factories, Sprout is built to work around people. It has a soft exterior, reduced pinch points, and animated “eyebrows” designed to make it seem approachable rather than threatening.
The company was founded by former Meta and Google engineers and had signed Disney and Boston Dynamics as early customers. Fauna’s roughly 50 employees, including founders Rob Cochran and Josh Merel, are joining Amazon’s Personal Robotics Group.
Fauna targets hotels, retail spaces, and research labs—environments where robots need to navigate around unpredictable humans. The modular software platform lets developers build applications without starting from scratch.
Amazon’s Robot History
Amazon isn’t new to robotics. The company acquired Kiva Systems for $775 million in 2012, renaming it Amazon Robotics. That unit now operates over one million robots across Amazon’s warehouse network.
But those are warehouse robots—machines that move goods in controlled environments. Getting robots to work in the real world is harder.
The Scout failure showed why. Sidewalk delivery robots looked promising in demos but couldn’t handle real neighborhoods. Curbs, stairs, dogs, construction, and unpredictable terrain broke the model.
RIVR and Fauna represent different approaches to the same problem: building robots that can actually function outside controlled warehouses.
What This Means
Two acquisitions in one week suggests Amazon is accelerating its robotics strategy. The company already deploys more robots than any other retailer. Adding quadrupeds and humanoids expands its capabilities into new domains.
For delivery, the near-term play is clear: RIVR robots help drivers move packages from vans to doorsteps, reducing the physical demands on workers and potentially speeding delivery times. Amazon said it would test this later in 2026.
For humanoids, the strategy is less obvious. Sprout is positioned for customer-facing applications—hotels, retail, entertainment. Amazon has talked about its Personal Robotics Group before, but this acquisition signals more serious ambitions in consumer-facing robots.
The bigger picture: Amazon is building toward a future where robots handle more physical work across its operations. Not just in warehouses, but in delivery vehicles, at customer doorsteps, and potentially in customer-facing roles.
The Competition
Amazon’s robot push comes as the humanoid market heats up. Tesla’s Optimus, Figure’s partnerships with BMW and OpenAI, and Agility Robotics’ work with Amazon’s logistics rivals all point toward a near-future where humanoid robots become commercially viable.
The RIVR acquisition puts Amazon ahead in quadruped delivery—a space with fewer competitors. The Fauna acquisition puts it into the humanoid race alongside much larger rivals.
Whether these robots eventually work well enough to justify the investment remains an open question. Amazon’s Scout failed. Many robotics startups have failed. But Amazon has the scale to absorb failures while continuing to push forward.
The message from this week: Amazon is not giving up on robots. It’s doubling down.