“Muscles never clamp directly onto bone.” This anatomical truth, as simple as it sounds, has become the blueprint for a leap forward in biohybrid robotics. For years, engineers have been able to grow muscle tissue in the lab and attach it to robotic frames, creating machines that move with lifelike smoothness. Yet the mechanical mismatch between soft muscle and rigid skeleton has stubbornly limited speed, strength, and durability.
Synthetic tendons are flexible hydrogel connectors developed by researchers at the Massachusetts Institute of Technology that bridge living muscle to mechanical frames with unprecedented efficiency. The idea is taken from vertebrate anatomy: tendons serve as a mechanical intermediary, tuned in stiffness to transfer force while protecting muscle from damage.
The team, led by Ritu Raman, assistant professor of mechanical engineering, fabricated narrow strips of engineered skeletal muscle from C2C12 mouse myoblasts, matured over two weeks into fibers responsive to optogenetic blue-light stimulation. At each end, they bonded a hydrogel tendon, just one millimeter of overlap, using a formulation pioneered by Xuanhe Zhao’s group. This hydrogel combines two polymer networks for toughness and elasticity, and is chemically treated to adhere tightly to biological tissue without harming cell viability.
The researchers modeled the system as three springs-muscle, tendon, and skeleton-each with unique stiffness-before building the physical prototype. Simulation showed that tendons need to be far stiffer than muscle to prevent energy losses. After some tuning, the fabricated tendons performed as predicted, thus allowing the muscle-tendon unit to move a mechanical joint in a near optimal way.
When fixed onto a robotic two-fingered gripper designed by Martin Culpepper, the results, indeed, were striking: the tendon-linked muscle pinched three times faster and delivered about 30 times more force than muscle attached directly to the frame. The system endured over 7,000 contraction cycles without failure, while the power-to-weight ratio jumped elevenfold-meaning far less muscle mass could achieve far greater output.
This modular “muscle-tendon unit” could easily be swapped into different kinds of robot architectures from micro-scale surgical to autonomously exploratory machines. “We are introducing the use of artificial tendons as interchangeable connectors between muscle actuators and robotic skeletons,” said Raman. “Our goal is to make muscle actuators easier to use, like off-the-shelf parts.”
The breakthrough builds on a decade of biohybrid robotics advances, where living tissue actuators have been combined with synthetic skeletons to create crawlers, swimmers, and grippers. Muscle offers unique advantages over traditional actuators: each cell is an independent force generator, tissues can strengthen with use, and they can self-repair after minor damage. But until now, without the right soft–rigid interface, much of that potential has gone to waste.
Interface engineering represents a critical frontier. In other areas, mechanical metamaterials and programmed micro-fiber weaving have been employed by investigators to bond soft elastomers to rigid substrates, attaining high debonding strength and durability under cyclic loads; in many ways, MIT’s hydrogel tendons accomplish biologically the same objective-bringing about a mechanical gradient between muscle and skeleton that improves force transmission and reduces stress concentrations.
Hydrogels themselves constitute a very versatile platform in biohybrid design. Due to tunable stiffness, biocompatibility, and the particular ability to mimic extracellular matrix, they serve perfectly for anchoring cells while resisting mechanical load. In this application, the stiffness of the tendon hydrogel was tuned to roughly triple that of the muscle tissue, which could ensure effective energy transfer without overstretching.
Speed tests underlined the advantage: direct-mounted muscles lost motion at higher stimulation rates, while tendon-linked units maintained performance up to four contractions per second, and could hold steady force at even higher rates. Fatigue still occurred in the muscle after prolonged activity, but the tendon links stayed sound. The implications are far-reaching. In medicine, muscle-powered tools could one day execute gentle in vivo manipulations, or deliver targeted therapies, exploiting the adaptability and self-healing of living tissue.
In exploration, biohybrid robots might enter hazardous or distant environments, perform minor damage repair autonomously, and survive for extended operations. In research, modular muscle-tendon units provide a controlled platform to study muscle mechanics and injury and recovery when integrated with an engineered system. By replicating nature’s muscle-tendon-bone architecture in a scalable, modular format, the team at MIT has redefined the performance envelope for biohybrid robots. The work not only addresses a long-standing mechanical mismatch but also opens a path toward living-machine systems that are stronger, faster, and more durable than ever before.