Robotics Weekly: Agility Robotics Goes Public, GM's Automation Pivot, and China's Humanoid Surge
This week's digest covers Agility Robotics' $2.5B IPO, GM's significant workforce reduction in favor of cobots at Factory Zero, and Morgan Stanley's aggressive forecast for Chinese humanoid shipments. We also explore breakthroughs in navigation chips from MIT, new autonomous trucking initiatives in California, and NASA's advanced rover testing.
Podcast В· 2 min
Agility Robotics to Go Public at $2.5B Valuation
Agility Robotics has announced plans to go public via a SPAC merger with Churchill Capital Corp XI, aiming for a $2.5 billion valuation. The company, which trades under the ticker $AGLT, is set to become the first pure-play humanoid robotics firm on U.S. markets, outpacing competitors like Figure AI in reaching Wall Street. The deal is expected to generate over $620 million in proceeds, with a significant $200 million contribution from a group led by Foxconn. The capital injection is earmarked for scaling production of the Digit v5, a next-generation humanoid designed for broader industrial applications beyond fenced-off cells. Agility reports over 65,000 hours of real-world operation for its robots, which are currently deployed at sites including Toyota and Schaeffler. While the company claims over $300 million in committed orders, filings reveal this is largely tied to a single three-year contract for 1,000 robots. This move marks a critical test for the humanoid sector. Scaling robots that operate alongside humans is notoriously capital-intensive, and the public market will now scrutinize whether Agility can transition from successful pilots to sustainable, large-scale production reality.
GM Replaces 1,000 Workers with 50 Cobots at Factory Zero
General Motors has implemented a significant automation shift at its flagship Factory Zero EV plant, laying off over 1,000 employees and installing approximately 50 collaborative robots from Fanuc. The company describes the move as a modernization effort intended to improve safety and operational efficiency by handling repetitive, physically demanding assembly tasks. The layoffs have drawn immediate backlash, with the local UAW union filing grievances and expressing strong opposition to the decision. Union leadership has warned that Factory Zero could serve as a blueprint for automated EV production across the broader automotive industry, raising concerns about the future of labor in highly automated facilities. This rollout offers an early look at the tension between productivity gains and labor stability in the next generation of EV manufacturing. As automakers continue to chase "software-defined" factory models, GM's ability to balance these technological upgrades with union pressure and political scrutiny will likely set a precedent for the rest of the industry.
Morgan Stanley Doubles China Humanoid Forecast
Morgan Stanley has significantly revised its 2026 forecast for Chinese humanoid shipments, now projecting 50,000 units—up from an initial January estimate of 14,000. The firm anticipates the Chinese humanoid market will reach $2 billion this year and grow to $15 billion by 2030, with annual shipments eventually hitting 446,000 units. The demand is underscored by a major State Grid procurement order, which includes 500 humanoids, 3,000 dual-arm robots, and 5,000 quadruped robots. China currently accounts for over 80% of global humanoid shipments, with domestic leaders like Agibot, Unitree, UBTECH, and Leju driving the momentum. This growth is supported by Beijing’s national priority on "embodied AI," which includes subsidies and favorable lending for startups. This rapid shift from lab to deployment highlights a growing gap between Chinese and U.S. competitors. While China’s top players each shipped over 5,000 units in 2025, U.S. counterparts like Figure AI and Tesla have moved significantly fewer, suggesting that state-backed industrial policy is accelerating China's path to scale.
MIT Develops Low-Power Navigation Chip for Tiny Robots
MIT researchers have unveiled the Gleanmer chip, a new hardware solution designed to enable tiny robots to navigate complex environments with minimal power consumption. The chip processes depth-camera data into real-time navigation maps while consuming only about 6 milliwatts, comparable to the power usage of an LED. Unlike traditional systems that rely on memory-intensive 3D voxel maps, Gleanmer represents obstacles using Gaussian ellipsoids, which compresses complex geometry into a much smaller memory footprint. Its companion algorithm, GMMap, processes depth images in a single pass, eliminating the need to store entire images or repeatedly revisit raw pixels. This technology addresses a major bottleneck for micro-robotics: the inability to perform sophisticated spatial awareness without draining batteries. By moving serious 3D navigation capabilities into low-power edge devices, Gleanmer could enable new applications for inspection drones, microbots, and AR glasses that must operate on extremely tight energy budgets.
NASA Tests Advanced 'Ernest' Rover Prototype
NASA’s Jet Propulsion Laboratory (JPL) is testing a new four-wheeled prototype rover named Ernest, designed to advance the capabilities of planetary exploration. The rover features an active suspension system that allows it to lift individual wheels and switch between gaits, including squirming and obstacle-climbing, offering a significant upgrade over the passive rocker-bogie suspension used on previous Mars missions. In a seven-day field test in the Colorado Desert in March 2026, Ernest covered 16 miles in 37 hours, demonstrating speeds up to six times faster than the Perseverance rover. Its autonomy was trained using reinforcement learning in a high-fidelity JPL simulator, which allowed for thousands of hours of virtual testing in a single weekend. While not yet mission-ready, the 4-foot-frame prototype is designed to overcome the terrain limits that have historically forced mission planners to route around inaccessible areas. JPL suggests that this level of speed and terrain-handling could enable "genuine science road trips" across the Moon and Mars in the future.