Whether they're dancing on two legs or scrambling over rough terrain on four, robots are gaining traction on social media and in everyday life. They're already rapidly evolving in terms of capabilities and size, but according to Penn State Assistant Professor of Mechanical Engineering Baxi Chong, they may be on the brink of something even better. Chong is one of several Penn State researchers capitalizing on unique biological features found in the living ecosystem to develop and expand the field of biorobotics.
One of the most obvious features to take advantage of, Chong said, is size.
"The size of the robot matters," said Chong, who is also a Huck Institutes of the Life Sciences co-hire. "There has already been a lot of research on macroscopic robots, which are classified as about one foot. "Microrobots, which measure around or less than one centimeter, are also a developed community. But what about the in-between?"
The in-between robots, described as mesoscale, are less studied in the broader field of robotics, Chong said, despite their benefits: They can both traverse confined spaces and navigate terrain obstacles, setting them apart from the big and tiny robots, which typically can only do one or the other.
"If we want to design robots to move and act like animals we can control, at different sizes and capable of accomplishing different tasks, what better place to find inspiration than from animals?" asked Chong, who joined the Penn State College of Engineering in fall of 2025.
His primary research focus is locomotion within mechanical intelligence, inspired by creatures with unconventional bodies, such as snakes, centipedes and lizards.
Chong joined a diverse team of researchers at Penn State using animals and insects as inspiration for robotic innovation, including Jean-Michel Mongeau, Shuman Family Early Career Professor of Mechanical Engineering, who primarily works with insects like fruit flies and cockroaches; Bo Cheng, associate professor of mechanical engineering, who studies hummingbirds and fish; Margaret Byron, Tretheway Early Career Professor of Mechanical Engineering, who studies the fluid dynamics of multimodal locomotion - or, how animals move in multiple ways, like swimming and flying.
The robot Cambrian explosion
Chong became interested in designing better robots after witnessing the 2008 Sichuan earthquake, which resulted in nearly 90,000 casualties and 18,000 missing persons.
"I saw thousands of firefighters, first responders and search and rescue teams get buried in the rubble, sacrificing their lives to save innocent people," Chong said. "What if robots could do this search and rescue instead?"
No single type of robot could navigate the unpredictable aftermaths from different types of disasters, Chong said, but robots adapted to different terrains and situations might be able to better assist.
"When you go to a zoo, you'll see animals of all different shapes and sizes dwelling in their habitats," Chong said. "If one were to build a robot zoo, there currently wouldn't be that level of diversity."
For Chong, this presents an opportunity.
"There was a period about 538.8 million years ago when biodiversity exploded, typically referred to as the Cambrian explosion," Chong said. "I would like to facilitate the robot Cambrian explosion."
Chong's team does full-stack engineering, where the same developer manages all aspects of a development process. For Chong, this process begins by observing the creatures he models in his lab, studying their movement and capturing it for robotic design. Visitors to his lab might often find live centipedes.
"I record centipede locomotion," he said. "We track every movement, every leg. In the past, students used a device that had buttons for each centipede leg and clicked to record the centipede positions as they tracked its movement. It took forever to collect data. Now, using artificial intelligence (AI), you can train the AI to track the centipede."
Potential in 'the sophistication of the cockroach antenna'
Much like Chong, Jean-Michel Mongeau is studying the unique movement intricacies of insects, like fruit flies and the American Cockroach, whose antenna has roughly 10,000 mechano-sensors per centimeter. Mongeau said the sophistication of the cockroach antenna is aspirational for those in robotics, and he believes in the possibility of using his research to embody artificial intelligence, rather than just housing it on a server.
"There is a belief that cognition is beginning to be solved by neural networks," Mongeau said. "The best chess-playing algorithms can beat the best chess players in the world. But walking to the board and moving pieces is something that a robot cannot do as well as a human toddler. Our work is to understand motor control - how your brain and body collaborate to move a body effectively."
Mongeau and his research partners used micro-computed tomography (micro-CT), a technique that produces high-resolution images at the microscale, to scan a cockroach antenna. Using this information, they have developed a computational model of the antenna to simulate how different forces impact it and to better understand how it processes touch information. Mongeau is also working with Kaushik Jayaram, associate professor of bioengineering at the Imperial College of London, to build a small-scale robotic antenna.
"Many insects are nocturnal, so they rely heavily on touch to explore," Mongeau said. "These creatures can process incredibly rich information from their sense of touch and can sense small features in complex environments. If we can harness that ability in robotics, it could enable a new class of touch sensors."
Fish robots that 'ride the flow'
Bo Cheng is another researcher studying animal movement, primarily the fluid mechanics involved with flight in birds and insects like fruit flies, as well as in how fish swim. Cheng not only analyzes the physical mechanics of swimming fish but also their perception abilities - how they respond to stimuli and changes in their environment, and how they navigate complex or difficult spaces.
Much of Cheng's research is in establishing alternative robot methodologies that can inhabit the natural environment more easily. Many underwater robots use jet propellers, which can be noisy and startle wildlife. He builds bio-inspired robots that mimic fish movements and puts them under different forces of pressure underwater. This allows him to assess how fish react to environmental changes, including how they can sometimes capitalize on pressure to expend less energy while swimming.
"We want to build a robot that can be biologically immersive and not disturb the environment," Cheng said. "These robots are quiet, can ride the flow and better navigate in unstructured environments."
He also studies hummingbird flight to understand how to potentially integrate flapping wings in flying robots, as opposed to the quadrotor approach already in use.
"Hummingbirds are the only birds that can sustain a hover for hours," Cheng said. "They can migrate across the Gulf of Mexico in one day, nonstop. Some fly around the coast, but some fly directly across the water. Imagine that - a little bird, flying alone across the sea."
Some hummingbirds can flap up to 100 times per second and can easily reject disturbances while flying.
"We would like to reverse engineer the muscle force pulling their weight while flying," Cheng said, who noted that construction of hummingbird robots hasn't begun, but that he is eager to get the work underway.
Multimodality: Poetry in locomotion
Margaret Byron runs the Environmental and Biological Fluid Mechanics (EBFM) Lab, where her team explores questions within the field of fluid dynamics, the physics of flow. One major topic they research is how animals swim, with specific focus on how organisms generate flow to propel through natural fluid environments like water or air. The primary aspects of fluid flow, she said, are inertia, which keeps an object moving, and viscosity, which slows an object down. An organism's size and speed have significant impacts on how they're able to move through an environment.
"If you're large or moving quickly, you don't have to think about viscosity as much because you have so much inertia," Byron said. "But if you're a bacterium, your environment is dominated by viscosity, so your strategies for swimming are different from those in an inertia-dominated environment. We look at how swimming strategies can change from a viscous to an inertial environment."
Byron and her team are studying multimodality in animal locomotion, meaning animals that are capable of more than one type of movement like walking, flying or swimming. Some insects she studies include backswimmers, water boatmen and diving beetles, all of which primarily live underwater, but also breathe air and are capable of flight. Byron and her team are especially interested in how fluid dynamics mediates the transition from one mode to another.
"If you have legs that you use for both walking and swimming, how does the usage differ?" Byron asked. "What changes do you make between those two locomotor modes? If you're wading through shallow water, how do you decide to transition to full swimming?"
She also researches ctenophores, or comb jellies, which might appear similar to jellyfish, as they are both gelatinous and live in the ocean, but they come from different phyla. Comb jellies have small flexible paddles, roughly a millimeter in size, which they use in coordination to propel themselves with what's called a metachronal wave.
"The flow between the paddles is unique," Byron said. "One paddle makes a vortex, another makes a vortex, and those two vortices combine and interact."
The behavior of animals like comb jellies is hard to control directly; they might not always act in ways conducive to research interests, Byron noted, but a robot inspired by a comb jelly can mimic the organism's biological system and allow for closer study in a more controlled environment. Byron said - and her colleagues agreed - that biology and engineering research have a reciprocal relationship, with both fields informing one another.
Nature-inspired robots to understand nature
Each of these robotics projects has their own practical applications, from disaster recovery to rehabilitation to neuroprosthetics to agriculture and far more, with significant potential for multiple applications, the researchers said. Chong's robots, for example, are uniquely sized and adaptable to terrains, making them ideal for navigating complex terrain, debris or confined spaces. Not only could they perform search and rescue, but they also have practical use in fields like plumbing and agriculture.
And beyond practical applications, these robots offer a deeper understanding of the natural world, for which these researchers all share a deep passion.
"I believe the team at Penn State is not just copying nature, we're developing deep understandings of how animals move, how they control their movements and how we can utilize those principles," Byron said. "There are incredible things that engineers can design and bring to life that have never existed in the world. We can come up with things that are creative, innovative and can improve the quality of life."
