RIVR’s journey began in the world of cutting-edge robotics research, rooted in one of Europe’s most renowned academic institutions. Founded in April 2023 as a spin-off from the Robotic Systems Lab at ETH Zurich, the company originally launched under the name Swiss-Mile. It emerged with a bold vision: to transform how autonomous systems move, navigate, and interact with the urban world—bridging AI with the physical complexity of real-world environments.
From the outset, Swiss-Mile aimed to redefine mobile robotics. The founding team—Marko Bjelonic, Lorenz Wellhausen, Giorgio Valsecchi, and Alexander Reske—leveraged years of foundational research and practical deployment experience from the Robotic Systems Lab at ETH Zurich, led by Professor Marco Hutter.
Their journey began well before the company's inception. In 2018, the team was part of pioneering work that deployed one of the first artificial neural networks on real legged robots, marking an early shift toward what is known today as Physical AI. This foundational breakthrough laid the groundwork for adaptable, intelligent robotics systems capable of performing complex physical tasks.
Swiss-Mile's development was deeply rooted in physics-based simulation environments, a foundation that led to a close collaboration with NVIDIA. This partnership culminated in significant announcements at NVIDIA's GTC conferences in 2022 and 2025, highlighting the synergy between simulation and real-world robotic applications.
The team's capabilities were further demonstrated as core contributors to CERBERUS, the winning team of the DARPA Subterranean Challenge, which the Washington Post has called “The Super Bowl of Robotics.”. This prestigious competition focused on deploying autonomous robots in complex, underground environments, underscoring the team's ability to bridge academic research with practical performance.
Each co-founder brought specialized expertise:
In August 2024, Swiss-Mile raised a $22 million seed round, marking a key milestone in its commercialization efforts. Led by Jeff Bezos through Bezos Expeditions and HSG, the round also included participation from the Amazon Industrial Innovation Fund and Armada Investment, with continued support from the existing investor Linear Capital.
This high-profile investment wasn’t just a financial boost—it was a strategic endorsement of Swiss-Mile’s unique approach to Physical AI. With support from partners at the forefront of global e-commerce and supply chain innovation, the company doubled down on its mission: to build the world’s most intelligent mobile robot for urban logistics—powered by General Physical AI.
In January 2025, Swiss-Mile rebranded to RIVR. The name reflects the company's evolution from a research-based venture into a full-scale robotics company focused on logistics and delivery. “RIVR” evokes motion, connectivity, and fluidity—qualities at the core of the company’s technology and the systems it’s helping to build.
This transition marked more than a name change. It signaled a broader vision for the role of robotics in society: not just delivering packages, but enabling a new era of Physical AI—intelligent machines that can navigate and interact with the physical world as intuitively as humans.
Today, RIVR is actively deploying its platform with major logistics and parcel carriers, helping them reimagine how goods move through cities—from depot to doorstep. With robots that can maneuver through traffic, climb stairs, and even interface with urban infrastructure, RIVR is pushing the boundaries of autonomous delivery.
As cities become denser and consumer expectations rise, the demand for intelligent, adaptable, and efficient delivery solutions is accelerating. At RIVR, we are building toward this future by advancing General Physical AI—intelligent machines that can move fluidly through dynamic, human-centered environments and autonomously complete real-world tasks.
Our robots are designed to operate in unstructured, urban settings where traditional automation falls short. By combining wheels for speed and legs for adaptability, they can climb stairs, cross curbs, and navigate narrow spaces—reaching the last-100-yards where conventional delivery vehicles can’t. Each deployment contributes to a growing pool of real-world training data, enabling continuous improvements through reinforcement learning, supervised learning, and real-time decision-making.
The market opportunity for General Physical AI is rapidly expanding. Autonomous logistics for food and parcel delivery alone is projected to reach $1–2 trillion by 2030, as cost-effective, AI-driven systems reshape consumer behavior (ARK Invest). Our focus on last-mile delivery unlocks critical edge intelligence and operational insight—essential building blocks for generalizable robotics.
This isn't just about better delivery. It’s about enabling robots to take on a wide range of physical tasks—across logistics, mobility, and urban infrastructure—and doing so in a way that is scalable, safe, and sustainable. As AI software converges with robotics hardware, the potential for General Physical AI shifts from transformative to exponential, with a $24+ trillion global revenue opportunity on the horizon.
At RIVR, we’re helping to define a new category of urban robotics for the cities of tomorrow giving 1 human, the power of 1000 through General Physical AI.
.png)