Robot Malfunctions and Grabs Female Student During School Sports Day Performance, Experts Cite Immature Technology

A humanoid robot grabbed a dancing student during a performance at Xi'an Eurasia University's sports meet on April 23. The robot company blamed signal interference from too many drones at the venue, causing the machine to malfunction. The robot, sponsored by an alumni enterprise, was returned after the performance without causing injuries—but the incident exposed serious gaps in how these machines handle real-world complexity.

Gao Huan, an expert from Chongqing Normal University, cut through the hype: this wasn't AI awakening, just motion control failure. Insufficient algorithm robustness, sensor vulnerability to interference, and inadequate safety redundancy are common flaws across current humanoid robots. In February 2026, a Zhiyuan robot lost control and crashed into tables at a Haidilao restaurant in California—same root cause, same unfinished tech being pushed into commercial settings too early.

Compare this to consumer drones a decade ago: frequent crashes and loss of control were the norm until years of iteration made DJI products stable in complex environments. Humanoid robots are at that early drone stage now—good for demos, far from daily reliability. Manufacturers racing to showcase capabilities are creating these failure scenarios by skipping the hard work of stress-testing in messy real-world conditions.

Three problems need solving before humanoid robots belong in crowds: electromagnetic interference resistance strong enough for chaotic signal environments, sensor fusion algorithms robust enough to handle unexpected inputs without losing control, and multi-layered safety mechanisms that trigger immediate shutdown when anomalies occur. Deploying machines in dense public spaces before addressing these fundamentals treats audiences like beta testers. Progress takes time—don't rush it at people's expense.