Imagine the agility of a mountain goat as it effortlessly scales near-vertical terrains, the protective instinct of an armadillo as it rolls into a ball, or the adaptability of creatures that can traverse any environment thrown their way.
Now, envision these extraordinary traits fused into a single robotic system, designed to conquer the most challenging landscapes with ease.
A team of researchers, led by Josie Hughes at the CREATE Lab in EPFL’s School of Engineering, has created a unique new robot that redefines mobility and adaptability.
The experts set out to engineer a robotic device with the adaptability of creatures in the animal kingdom. They aimed to push the boundaries of robotic mobility by designing a machine that could transition seamlessly between different environments with minimal effort.
The goal was to develop a robot capable of navigating diverse terrains by dynamically changing its form as needed, just like nature’s most versatile animals.
After much research and development, the researchers developed GOAT – Good Over All Terrains – a promising new development in robotic locomotion and control.
This robust machine is capable of morphing between a flat, “rover” shape and a sphere. This flexibility allows the robot to switch between modes of mobility such as driving, rolling, and even swimming. The new development combines clever design with a surprising amount of energy efficiency.
“While most robots compute the shortest path from A to B, GOAT considers the travel modality as well as the path to be taken,” said Hughes.
“For example, instead of going around an obstacle like a stream, GOAT can swim straight through. If its path is hilly, it can passively roll downhill as a sphere to save both time and energy, and then actively drive as a rover when rolling is no longer beneficial.”
The team at the CREATE Lab developed a design that thrives in unpredictable environments.
The researchers drew inspiration from some of nature’s most agile and resilient creatures – kangaroos, with their powerful leaps, spiders with their precise locomotion, octopuses with their fluid adaptability, and cockroaches with their near-indestructible resilience.
This bio-inspired compliance allows GOAT to adjust its structural properties seamlessly, shifting from a flexible and maneuverable “rover” mode to a robust, impact-resistant sphere when necessary.
Constructed of low-cost materials, GOAT’s simple frame is made up of two intersecting, elastic, fiberglass rods, and four motorized, rimless wheels.
The robot’s transformation is made possible by two winch-driven cables that shorten like tendons to draw it tightly into a ball. This protects the onboard computer, sensors, and battery in its core – much like a hedgehog protects its underbelly.
Compliance is not just about physical adaptability; it’s also about navigation. Max Polzin, a PhD student at the CREATE Lab, noted that GOAT’s compliance allows it to navigate effectively with minimum sensing equipment.
Lacking cameras, the robot relies solely on a satellite navigation system and an inertial measurement unit to measure its orientation.
“Most robots that navigate extreme terrain have lots of sensors to determine the state of each motor but, thanks to its ability to leverage its own compliance, GOAT doesn’t need complex sensing,” said Polzin.
“It can leverage the environment, even with very limited knowledge of it, to find the best path: the path of least resistance.”
Future research will explore advanced algorithms to better exploit the unique capabilities of morphing, compliant robots like GOAT. The potential applications vary from environmental monitoring to disaster response and even extraterrestrial exploration.
“Robots like GOAT could be deployed quickly into uncharted terrain, with minimal perception and planning systems, allowing them to turn environmental challenges into computational assets,” said Hughes.
“By harnessing a combination of active reconfiguration and passive adaptation, the next generation of compliant robots might even surpass nature’s versatility.”
With continued advances in materials science, AI-driven decision-making, and energy-efficient actuation, these robots could become indispensable tools for navigating environments that are too hazardous or unpredictable for traditional machines.
The full study was published in the journal Science Robotics.
Image Credit: © CREATE EPFL
—–
Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.
Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.
—–