This Company is Transforming Industrial Robots to be More ‘Human-Like.’

Early perceptions of robots portrayed their movements as stiff and jerky, an idea reflected in the “robot dance” craze of the 1980s.

Over time, robots have developed smoother and more natural movements, especially humanoid robots, though they represent a small fraction compared to the vast number of industrial robots that have been assembling products like cars for years.

Currently, approximately 3 million robots operate in factories globally, with a third employed in the automotive sector, as per industry data. Micropsi Industries aims to make industrial robots more human-like. Founder Ronnie Vuine explains, “Our control system enables industrial robots to perform tasks that would otherwise be impossible, providing them with hand-eye coordination and the ability to adapt to changing factory conditions.” Their flagship product, MIRAI, leverages AI and cameras to train robots for tasks that go beyond traditional pre-programmed operations.

Micropsi equips robots with the ability to adjust to changing conditions.

Ronnie Vuine’s interest in AI sparked during his studies at Berlin’s Humboldt University in the 2000s. His research group explored how machines could learn autonomously in real-world environments without an engineer’s guidance—essentially figuring out how to adapt and function independently.

At that time, AI was not widely embraced, but Google’s acquisition of Deep Mind in 2014 marked a turning point, signaling AI’s mainstream potential. This motivated Vuine and his team to establish Micropsi Industries the same year.

Micropsi now focuses on enhancing manufacturing robots for various brands. According to Vuine, the automotive industry leads in large-scale robot deployment due to the complexity and scale of car production. While planes are more intricate, their production volume is much smaller compared to cars, which showcase the pinnacle of automation.

Micropsi’s MIRAI system demonstrates this capability by enabling a robotic arm to perform delicate tasks, like picking up a dangling computer cable and plugging it into a switch—an operation too intricate for traditional programming. Vuine explains that MIRAI trains the robot to master such tasks in about an hour with human guidance. Once trained, the robot uses cameras and lighting to complete the task autonomously, adapting to unpredictable movements like the cable’s jiggle with consistent accuracy.

Micropsi’s system enables the robot to grasp a small, dangling cable and insert it into a switch.

Micropsi’s system opens new possibilities for automating tasks traditionally handled by humans, particularly in electric vehicle production. “With the shift to electric cars, there are many more cables to connect,” explains Vuine. “This is also critical in electronics, where ribbon cables link circuit boards. Tasks like these couldn’t previously be automated—they required human labor or product redesign for manufacturability.”

Having recently relocated its headquarters from Berlin to San Francisco, Micropsi aims to expand its focus beyond automotive to include products like power tools, household appliances, and entirely new sectors, such as logistics. The system may even power humanoid robots in the future. Vuine notes, “Our software could easily apply outside factories, such as in service robots that wash dishes. We’ve even done playful demos showcasing these capabilities.”

The main challenge for broader applications lies in the robots themselves, not the software. Vuine highlights, “Robots are made of rigid metal, unlike humans. If they hit you, it’s painful, so they must operate slowly with extensive safety measures. This results in machines that are too costly and impractical for home use. We haven’t solved that issue yet.”