Penn’s knotted robot jumps nearly two meters without a motor
A millimeter-scale knot releases stored elastic energy without a motor or battery.📷 AI-generated / Tech&Space
- ★The robot uses a Kevlar core and LCE shell in a fiber about one millimeter thick
- ★The knot acts as a latch: at 60 to 90 C it releases elastic energy into jumping, spinning, or gliding
- ★Maple-seed wings help control flight and soil penetration, but activation temperature and field robustness remain open
Penn Engineering's robot looks like a string trick until you see what is actually being programmed. According to TechXplore, Shu Yang and Yaoye Hong's team uses a fiber about a millimeter thick: a Kevlar core provides stiffness and energy storage, while a liquid crystal elastomer shell responds to heat.
The knot is not a tie here. It is a latch. When temperature rises to about 60 to 90 C, the LCE shell contracts and untwists, the knot loosens, and stored elastic energy becomes kinetic energy. The result is a jump of nearly two meters, hundreds of times the robot's own size. This is not a motor in a small package. It is geometry behaving like an actuator.
The best detail is that topology changes behavior. An overhand knot can create a flip, a figure-eight can spin like a propeller, and more complex knots release energy in phases. The Science paper summary describes programmable millimeter-scale soft robots built from Kevlar-LCE fibers, with explosive energy release on heating. In other words, the code is not in a microcontroller. It is in how the fiber is tied.
A Kevlar-LCE fiber turns knot topology into a programmed actuator, but fields, humidity, and heat triggers are not a lab floor.
Knot geometry acts as the robot's mechanical program for different motions.📷 AI-generated / Tech&Space
A wing inspired by maple seeds turns the jump into more controlled flight or a return path. The same principle can help seed planting because the robot gets a vertical impulse for soil penetration. TechXplore also cites higher penetration pressure than earlier rain-activated seed carriers, which explains why the researchers are looking toward agriculture at all.
But a laboratory flight and field deployment are not the same league. Activation at 60 to 90 C requires a controlled heat source or an environment much tidier than a real field. Humidity, dust, wind, shade, uneven soil, and organic debris can change the timing of release or the direction of impact. A robot without a battery is elegant only if activation does not become hidden infrastructure.
Materials are not free across cycles either. Kevlar adds robustness, but the LCE shell has to survive repeated heating, bending, and soil contact. The source mentions future work on more environmentally friendly materials and lower activation temperature, which is the right direction. For disposable planting, cost and degradability may matter as much as jump height.
That makes the biggest result mechanical, not agricultural. Penn has shown that a knot can be memory, latch, and actuator at the same time. If future versions lower activation temperature and survive dirty terrain, this could become a useful class of electronics-free microrobots. For now, the honest reading is that soft robotics has found a way to program matter as seriously as traditional robotics programs motors.
