Coco Robotics Launches New Physical AI Research Lab Led by UCLA Professor Bolei Zhou
Coco Robotics, the Los Angeles-based startup best known for its autonomous last-mile delivery robots, is turning five years of real-world data into its next big innovation push. The company has launched a new Physical AI Research Lab, led by University of California, Los Angeles (UCLA) professor Bolei Zhou, one of the most recognized experts in computer vision and robotics.
Announced Tuesday, the initiative also marks Zhou’s appointment as Coco Robotics’ Chief AI Scientist — signaling the company’s commitment to advancing AI-driven automation in robotics.
When Coco Robotics launched in 2020, its delivery bots were guided by teleoperators who helped them navigate challenging urban environments. CEO and co-founder Zach Rash told TechCrunch that the long-term vision has always been full autonomy to drastically reduce last-mile delivery costs. Now, after years of collecting data from real-world deliveries, the company is ready to take automation to the next level.
“We’ve collected millions of miles of data across the most complex city environments, and that’s incredibly valuable for training robust, real-world AI systems,” Rash said. “We now have the scale to accelerate groundbreaking research in physical AI.”
According to Rash, bringing Zhou on board was an obvious choice. Zhou’s academic research at UCLA has focused heavily on robotic navigation, reinforcement learning, and micromobility systems — all highly relevant to Coco’s mission.
Both Rash and co-founder Brad Squicciarini, who are UCLA alumni, have collaborated with Zhou before, even donating one of their delivery robots to support the university’s robotics program.
“He’s among the leading researchers worldwide in robotic navigation and physical AI,” Rash said. “He’s already helping us attract top researchers to join Coco and accelerate our innovation pipeline.”
A New Chapter for Robotics Research
The Physical AI Lab operates independently from Coco’s existing collaboration with OpenAI, which allows the company to leverage OpenAI’s large language models while contributing data from its robot fleet for AI research.
For now, Coco Robotics plans to keep its research findings internal — not to sell the data, but to enhance automation, optimize efficiency, and improve its proprietary local AI models. Rash emphasized that the insights will also help cities identify and fix infrastructure issues that impact delivery robot performance.
“Our goal is to deliver higher-quality service at much lower costs,” Rash said. “By improving affordability for both businesses and customers, we can drive massive growth across the delivery ecosystem.”
The new lab represents a strategic step toward advancing physical AI, combining robotics data, academic expertise, and real-world application — and further positioning Coco Robotics as a key innovator in the autonomous delivery industry.
TechCrunch Event
San Francisco | October 27–29, 2025
What did you think of this news? Leave a comment below and share it on your social media to help more people stay informed about the latest in technology, innovation, science, and robotics!
This news was originally published in:
Original source
