Coco Robotics Unveils Fully Autonomous Coco 2 Platform to Scale Urban Delivery Economics

New updates have been reported about Coco Robotics.

Coco Robotics has launched Coco 2, a next-generation autonomous delivery robot platform that shifts the company from human-guided operations to full autonomy, with direct implications for unit economics, scalability, and city expansion. Trained on millions of miles in diverse conditions across the U.S. and Europe, Coco 2 is designed to operate on sidewalks, bike lanes, and permitted roads, broadening the company’s addressable logistics footprint well beyond traditional sidewalk food delivery.

For executives evaluating Coco as a partner or competitor, the new platform promises up to 50 percent faster delivery times versus the prior generation, roughly tripling uptime, and improving resilience to weather and wear, all of which increase daily order capacity and reduce cost per mile. Coco positions Coco 2 as a general-purpose urban logistics asset that can move goods for restaurants, grocers, pharmacies, and local retailers, and it already powers autonomous delivery for platforms such as Uber Eats, DoorDash, and Wolt while serving more than 3,000 merchant locations. The company, which has completed over 500,000 zero-emission deliveries, plans to scale its fleet to thousands of robots globally by year-end, leveraging what it calls the industry’s largest dataset of sidewalk-robot operations to accelerate deployment in new cities.

Coco 2’s autonomy stack is built on NVIDIA’s robotics ecosystem, including Omniverse-based simulation with Isaac Sim and Isaac Lab, plus NVIDIA Cosmos world foundation models to generate synthetic street scenarios that allow the robots’ AI to train on edge cases before field deployment. Onboard, Coco 2 runs on NVIDIA Jetson Orin NX, enabling high-performance edge processing without constant cloud connectivity, which supports real-time perception, route planning, and maneuvering in dense urban environments. Coco’s leadership frames this as the arrival of “physical AI” at scale, using a continuous feedback loop between millions of real-world miles and advanced simulation to make the fleet more adaptive and efficient over time. For stakeholders, the launch signals Coco’s intent to be a core infrastructure provider for last-mile urban logistics, with a strategy centered on cost-efficient autonomy, rapid multi-city rollout, and platform integration with major delivery networks.