Unlocking Mobility: Artificial Muscles Revolutionize Robotic Leg Movement
Researchers at ETH Zurich and the Max Planck Institute for Intelligent Systems have developed a groundbreaking robotic leg powered by artificial muscles known as HASELs (Hydraulically Amplified Self-healing Electrostatic actuators). Inspired by living organisms, this innovative leg can jump and adapt to various terrains in an agile and energy-efficient manner, marking a significant advancement in the field of robotics.
A Leap Forward in Robotics
For nearly 70 years, robotics has predominantly relied on electric motors—a technology dating back over two centuries—to drive mechanical movements. This reliance has limited robots’ mobility and adaptability, making it challenging for them to emulate the fluid and versatile movements of living creatures. Traditional robotic systems often struggle with agility and energy efficiency, particularly when navigating uneven terrains.
“The field of robotics is making rapid progress with advanced controls and machine learning,” explains Christoph Keplinger, one of the leading researchers on the project, “but there has been much less progress with robotic hardware, which is equally important.” With the introduction of this new muscle-powered robotic leg, researchers are presenting hardware concepts that hold the potential for disruptive innovation in robotics.
Mechanism of Action: Electrostatic Muscles
The artificial muscles, HASELs, function on an electrostatic principle similar to static electric interactions experienced when rubbing a balloon on hair. These actuators are oil-filled plastic bags, akin to those used for making ice cubes, with electrodes coated on half of each bag’s surfaces. When a voltage is applied, the electrodes are attracted to each other due to static electricity, causing the bag to shorten as the oil is displaced to one side.
“As soon as we apply a voltage to the electrodes, they are attracted to each other due to static electricity,” explains doctoral student Thomas Buchner. “Similarly, when I rub a balloon against my head, my hair sticks to the balloon due to the same static electricity.” By increasing the voltage, the actuators contract, mimicking the action of biological muscles.
Pairs of these actuators are attached to a skeletal structure by tendons, replicating the paired muscle movements found in living organisms: as one muscle contracts, its counterpart extends. The researchers use computer code to control which actuators contract and which extend, enabling precise movements without the need for complex sensors.
Energy Efficiency Reinvented
Compared to conventional robotic legs powered by electric motors, the new system is remarkably more energy-efficient. Electric motors often generate excessive heat, leading to energy loss and necessitating additional hardware for heat management, such as fans or heat sinks. In contrast, the electro-hydraulic design of the artificial muscles maintains a consistent temperature, eliminating unnecessary energy consumption.
“On the infrared image, it’s easy to see that the motorized leg consumes much more energy if, say, it has to hold a bent position,” Buchner notes. “The temperature in the electro-hydraulic leg remains the same. Typically, electric motor-driven robots need heat management, which requires additional hardware.” Fellow researcher Toshihiko Fukushima adds, “Our system doesn’t require them.”
Navigation and Agility: Adapting to Uneven Terrain
One of the remarkable capabilities of the robotic leg powered by artificial muscles is its ability to jump and adapt adeptly to various terrains, including rocky, sandy, or soft surfaces. The leg’s musculoskeletal system possesses sufficient elasticity to adapt flexibly to the terrain in real time, much like human and animal legs.
“If we can’t bend our knees, for example, walking on an uneven surface becomes much more difficult,” explains Robert Katzschmann, head of the Soft Robotics Lab at ETH Zurich. “Just think of taking a step down from the pavement onto the road.” The artificial muscles allow the robotic leg to adjust its joint angles upon landing, depending on whether the surface is hard or soft, without the need for complex sensor systems.
In contrast to electric motors, which require sensors to constantly monitor the leg’s position and adjust accordingly, the artificial muscle system reacts instinctively to environmental changes. This adaptability is driven by just two input signals: one to bend the joint and one to extend it.
Practical Applications and Future Possibilities
The implications of this research are far-reaching. The adaptability and agility afforded by artificial muscles open up numerous applications—from robotic systems in logistics and product handling to assistive robots capable of performing household tasks. The technology holds particular promise for soft robotics, where movements need to be highly customized and responsive.
“Electro-hydraulic actuators are unlikely to be used in heavy machinery on construction sites,” Katzschmann acknowledges, “but they do offer specific advantages over standard electric motors.” This is especially evident in applications such as robotic grippers, where the movements must be finely tuned depending on whether the object being gripped is a ball, an egg, or a tomato.
Researchers envision integrating these robotic legs into quadrupedal or bipedal robots, potentially allowing for their deployment in rescue missions or sectors requiring high agility and safety. “If we combine the robotic leg in a quadruped robot or a humanoid robot with two legs, maybe one day, when it is battery-powered, we can deploy it as a rescue robot,” Katzschmann elaborates.
While the current iteration of the robotic leg is tethered and restricted to circular jumps, future work aims to overcome these limitations, opening the door to developing real walking robots with artificial muscles.
A Collaborative Effort: The Max Planck ETH Center for Learning Systems
This groundbreaking research is the result of a collaborative partnership between ETH Zurich and the Max Planck Institute for Intelligent Systems under the Max Planck ETH Center for Learning Systems (CLS). Established in 2015, CLS addresses interdisciplinary research questions in the design and analysis of natural and man-made learning systems. The core element of this partnership is a co-advised doctoral fellowship program, fostering future research leaders in the field.
The study represents an ideal example of collaborative research on physical intelligence under the CLS umbrella. Each doctoral fellow has a supervisor from both institutions and spends time at each location, promoting a rich exchange of ideas and expertise.
Conclusion
Artificial muscles in robotics represent a pivotal shift in the capability and functionality of robotic systems. By mimicking the mechanics of biological muscles, the new electro-hydraulic actuators offer enhanced energy efficiency, agility, and adaptability. This technology opens new horizons in soft robotics, with potential applications ranging from assistive devices to agile rescue robots.
As the field of electrohydraulic actuators is still young, the researchers anticipate that continued innovation and mass production of components will further advance the technology, making it more accessible and affordable. The fusion of advanced hardware concepts with cutting-edge controls and machine learning heralds a new era in robotics, bringing machines one step closer to the adaptability and finesse of living organisms.
Buchner TJK, Fukushima T, Kazemipour A, Gravert SD, Prairie M, Romanescu P, Arm P, Zhang Y, Wang X, Zhang SL, Walter J, Keplinger C, Katzschmann RK. “Electrohydraulic musculoskeletal robotic leg for agile, adaptive, yet energy-efficient locomotion.” Nature Communications.s these innovations will undoubtedly shape the future of robotics and AI applications.
Post Comment