Coco Robotics has launched Coco 2, its next-generation autonomous delivery robot platform, marking a significant milestone in the evolution of urban logistics. The new system signals the company’s transition from human-guided robotics to fully autonomous operations powered by physical AI and real-world data.
From Human-in-the-Loop to Full Autonomy
Coco 2 is trained on millions of real-world operational miles across dense and unpredictable city environments, including Los Angeles, Miami, and Chicago. The platform leverages extensive deployment data to enhance navigation precision, adaptability, and safety.
According to Zach Rash, CEO and Co-Founder of Coco Robotics, every operational mile strengthens the fleet’s collective intelligence, enabling real-time adaptability in new cities.
Operational Expansion: Beyond Sidewalk Delivery
Originally focused on sidewalk-based food delivery, Coco 2 now expands into bike lanes and roads where permitted. The company reports up to a 50% reduction in delivery times compared to its previous generation, alongside three times longer uptime and improved weather resilience.
The platform supports a broader range of merchants — from restaurants and grocers to pharmacies and retailers — and integrates with platforms such as Uber Eats, DoorDash, and Wolt. Coco currently serves more than 3,000 merchants and has completed over 500,000 zero-emission deliveries across the U.S. and Europe.
Technology Stack: NVIDIA-Powered Intelligence
Coco 2’s intelligence is built using simulation and AI tools from NVIDIA, including Omniverse libraries, Isaac Sim, Isaac Lab, and Cosmos world foundation models. The robots run on NVIDIA Jetson Orin NX hardware, enabling high-performance edge computing without reliance on cloud connectivity.
This integrated AI stack allows the robots to simulate real-world urban scenarios, anticipate movement, plan routes, and execute maneuvers safely before live deployment.
Share your Details for subscribe