Exploring the Mysteries of Motion

Roboticists often turn to the natural world for inspiration, especially when it comes to motion. Functional prototypes such as butterflies and rattlesnakes already exist in nature, so the challenge becomes understanding the intricacies of their motions well enough to advance robotics design. We discuss two university labs that seek inspiration in nature to design robots that can walk and jump and robotic limbs that help restore normal motion for amputees.

The latest biolocomotion project at Johns Hopkins University’s Flow Physics and Computation Lab involves spider crickets, which can leap a distance that is about 60 times their body length and still land on their feet. For humans, this feat would be equivalent to jumping the length of a football field.

Director Rajat Mittal and his students spent eight months using high-speed videography (400 frames per second) to study spider crickets in motion to discover how they jump so far. The film shows that, during the initial part of a jump, the crickets assume a streamlined posture that reduces drag and helps them to jump farther.

“However, shortly before reaching the apex of their jump, the crickets rapidly transition to a ‘stabilization posture,’” Mittal said. “They extend all their limbs outwards and their bodies assume a nearly constant angle from the horizontal of 55 to 60 degrees.”

Initial analysis suggests that this body posture is stable enough to prevent angular rotations and allows the crickets to maintain aerodynamic stability. By maintaining this orientation, the crickets land first on their powerful hind limbs, allowing them to “stick” the landing.

One application for this knowledge is the development of small jumping robots that can traverse rugged terrains, search for victims after earthquakes, or carry out other dangerous tasks.

“One of the things I enjoy the most about our research is the chance to discover beauty in the seemingly mundane world around us,” said Mittal. “Everybody considers these insects as nothing more than pests, but once we start looking at them closely we find this amazing beauty and grace that evolution has imbued in them.”

Tethered Prosthesis

More than one million Americans have had a leg amputation. Many individuals walking with prostheses fear falling, and no wonder: Human limbs evolved to walk over uneven terrain, and recover instantaneously from trips and shoves. Artificial limbs do not come close.

Hartmut Geyer, director of the Legged Systems Group at Carnegie Mellon, hopes to close the gap between the natural and the artificial by developing robotic limbs that mimic human reflexes and dexterity.

“The modeling suggests that reflex-like control of the knee joint would enable it to recover from trips and other disturbances that might otherwise lead to a stumble or fall,” said Steve Collins, associate professor of mechanical engineering and robotics, who works closely with Geyer on wearable robots. “My laboratory is developing a prosthetic knee emulator system to test these strategies with above-knee amputees in laboratory experiments.”

Based on prior emulators, Collins expects the tethered prosthesis his team develops will be exceptionally responsive and lightweight, allowing for very precise testing. The team will also develop prosthesis and exoskeleton emulators for other joints and incorporate additional degrees of actuation. Their near-term goal is to create a complete lower-limb system.

Geyer hopes his control strategies will lead to the development of advanced prosthetics to help above-knee amputees. Their lack of knee joints prevents them from shifting their weight quickly when they stumble. Robotic systems with control systems that provide very rapid feedback could help them maintain balance and control as the walk, even over uneven ground.

“Being involved in robotics research to help improve the quality of life for people who depend on technology to assist, replace, or restore their physical capabilities is a truly rewarding experience,” said Geyer. “It is also an exciting area of research for mechanical engineers who have a strong interest in multidisciplinary approaches to solving engineering challenges.”

Mark Crawford is an independent technical writer.