Robots Get Real Living Muscles That Jump and Swim 10x Better

Trading electric motors for lab-grown tissue

Instead of turning to miniature electric motors or rigid pneumatics, the Soft Robotics Lab at ETH Zurich grew real living muscle tissue to drive mechanical motion.

Traditional micro-robotics constantly runs into the same bottleneck: tiny electromagnetic actuators are stiff, heavy, and power-hungry. To sidestep those limits, engineers at ETH Zurich cultured living mouse muscle cells onto synthetic scaffolds, creating centimeter-scale bioactuators that contract like natural biological tissue.

While industries face supply constraints like the SK Hynix memory shortage impacting robotics and hardware supply chains, biohybrid research looks past conventional component bins entirely. Earlier attempts at tissue-driven machines produced contractions so faint that researchers had to measure them under microscopes. Co-lead author Aiste Balciunaite noted that the team discarded fixed muscle templates entirely, shifting to an iterative computer-designed architecture. The result is a system that yields 10 times the mechanical motion of prior iterations, producing movement easily seen with the naked eye.

ETH Zurich biohybrid robot specs and performance

How biohybrid robots jump, swim, and walk

The performance gain did not come from piling on more flesh. Instead, the engineers discovered that the mechanical interface between tissue and synthetic bone dictated overall strength. By molding specialized anchoring grooves directly into the artificial scaffolds, the contracting fibers secured a firmer grip, transferring force far more efficiently.

That grip translates into diverse forms of locomotion. The researchers demonstrated swimming units, grasping grippers, and walking variants that trudge forward at a deliberate 6.2 millimeters per second. Because living fibers retain biological variations, each actuator behaves with slight organic differences—a marked contrast to the identical tolerances of mass-produced silicon and steel.

The life support catch

Nobody is building an autonomous bio-terminator just yet. The fundamental drawback of biological hardware is that it requires biological maintenance. Today, these robots can only operate submerged inside a nutrient-rich bath that feeds the cells and keeps the muscular tissue alive.

Moving these machines out of the lab dish will require miniature circulatory systems or protective synthetic membranes that supply metabolic fuel. Until material scientists solve the desiccation problem, living actuators will remain strictly aquatic.

Availability and real-world deployment

This remains an active academic research project and is not a commercial product available for purchase. Practical deployment in medical implants or micro-exploration tools across the US and Europe depends on overcoming cellular lifespan hurdles outside controlled liquid environments.

By treating living cells as functional mechanical components rather than raw biological curiosities, ETH Zurich has pushed soft robotics closer to the agility of natural organisms. Solving cellular fuel delivery is the next hard frontier.