Humanoid Robots Excel at Entertainment, Struggle with Factory Work

While humanoid robots can be hired for entertainment purposes in Beijing, such as dancing at product launches or participating in marriage proposals, their integration into regular factory jobs faces significant challenges. Tens of thousands of these robots are expected to be produced this year, primarily in the Far East, but only a small fraction are anticipated to perform consistent industrial tasks. The main obstacles are physical limitations and the variability of real-world environments. A robot that performs flawlessly on a stage might struggle in a factory with uneven floors, unexpected objects, or human workers entering its path. Furthermore, factories must assess the robots' reliability for extended shifts, their safety around humans, and their ability to adapt to changing daily conditions.

A pilot program by BMW demonstrated the potential for humanoid robots in manufacturing. For ten months, the car manufacturer utilized a Figure robot for ten-hour shifts, five days a week. The robot handled over 90,000 metal components, walked 1.2 million steps, and contributed to the production of more than 30,000 vehicles. This successful integration was not accidental; it involved close collaboration between IT, safety, process management, and logistics teams from the outset. The robot was connected to existing systems and placed in a relatively automated zone, highlighting that success depended on the factory's ability to create a safe and stable working environment around the robot.

The economic viability of using humanoid robots in factories hinges on their operational hours. Longer working periods make the initial investment more justifiable. However, lengthy testing phases, restricted work areas, and narrow task definitions often leave robots idle, increasing the cost per working hour. This caution is warranted, as a robot's error in a factory setting, unlike a chatbot's mistake, can lead to severe damage or endanger human lives. Employers and regulators need to thoroughly understand robot behavior and responses to changing conditions before assigning them more critical tasks.

Information sharing also impedes the adoption of humanoid robots in industry. Knowledge gained from successes, failures, and challenges within one company is rarely transferred to others. Even when shared, the specific production line speed, product types, proximity to human workers, and physical conditions vary significantly between companies, necessitating individual assessments for each robot's deployment and human intervention needs. Much of this crucial knowledge remains confined to pilot project data.

Dr. Tal Cohen, a General Partner at NextGear Ventures, notes that the real difficulty arises after initial demonstrations. While startups excel at building the technology and factories at operating production lines, bridging the gap into a reliable working method requires dedicated effort. He sees an opportunity for Israeli companies in this area, despite robot manufacturing likely not occurring in Israel. Israel's proven strengths in systematic data monitoring, cybersecurity, sensor systems, and enterprise system integration can help factories monitor robot performance, detect environmental changes, and determine when human intervention is necessary. This data-driven approach can enable robots to work longer hours and perform more complex tasks. The true measure of success for these robots, and the associated Israeli opportunity, will be their ability to transition from demonstration to consistent, long-term production floor work.