Coco Roboticsstartup known for its fleet of last-mile delivery bots wants to extract more information from five years of knowledge collected by its robots. The answer: a physical artificial intelligence lab run by a professor at the University of California, Los Angeles (UCLA). Bolei Zhou in the lead.
Coco Robotics, which made the announcement on Tuesday, said Zhou had also joined the Los Angeles-based startup as its chief artificial intelligence scientist.
When the company launched in 2020, it used teleoperators to help bots overcome obstacles on delivery routes. Coco Robotics co-founder and CEO Zach Rash told TechCrunch that the company’s goal has at all times been to operate last-mile delivery robots autonomously to lower overall delivery costs. Now Rash said the company has collected enough data to delve deeper into automation.
“We have millions of kilometers of collected data in the most complex urban environments possible, and this data is extremely important in training any useful and reliable real-world AI systems,” Rash said. “We’re at the point now where we have enough data scale that I think we can really accelerate a lot of the research that’s going on around physical AI.”
The decision to appoint Zhou as the leader of this motion was “obvious,” Rash said. Rash said Zhou’s research in computer vision and robotics largely focuses on micromobility, as opposed to full-size vehicles.
Coco Robotics has also worked with Zhou before. Both Rash and his co-founder Brad Squicciarini are UCLA graduates and have even donated one of their bots to the school’s research lab.
“[Zhou] is one of the world’s leading researchers in robotic navigation, reinforcement learning, and many technologies and research areas that are very relevant to us,” Rash said. “He has already been able to recruit some of the world’s leading researchers that he has worked with in the past to join Coco and help accelerate the work on our end.”
Techcrunch event
San Francisco
|
October 27-29, 2025
This new research lab is different from the robotics startup’s collaboration with OpenAI, which allows Coco Robotics to use OpenAI models while the AI research lab accesses data collected by the company using robots.
For now, Coco Robotics plans to use the information and research collected in the laboratory for its own purposes. Rash said the company has no plans to sell the data to other firms.
Rather, it would serve the company to improve automation and efficiency, which is able to mainly apply to the local models on which its robots run. Rash said additionally they plan to share the results of their research with the cities where they operate, if possible, to help remove obstacles and infrastructure that decelerate their bots.
“The success of this lab really depends on us offering a higher quality service at an extremely low price,” Rash said. “How do we lower our costs? How do we make it much more affordable for businesses and customers? I think it will drive tremendous growth in this ecosystem.”
