When people think of robots, they often imagine machines that look like humans, dogs, or insects.
For years, engineers have designed robots by copying the shapes and movements found in nature.
But researchers at Duke University believe there may be a better approach.
Instead of focusing on what a robot looks like, they argue that the key is designing a robot that can move equally well in every direction.
Using this idea, they created a remarkable new robot called Argus, a machine that looks more like a sea urchin than anything seen in science fiction.
The robot is described in the journal Science Robotics.
Argus has no front, back, left, or right side. It consists of a central body surrounded by 20 telescoping legs that extend outward in all directions. Each leg is equipped with a depth-sensing camera, allowing the robot to see its surroundings from nearly every angle.
The unusual design gives Argus abilities that are difficult for conventional robots to match. During testing, it successfully rolled across grass, sand, wet surfaces, tree roots, and dense vegetation. It could climb over obstacles, carry heavy loads, and continue operating even after several of its legs were damaged.
Researchers say the robot’s performance comes from a design principle they call “dynamic isotropy.” This concept measures how evenly a robot can accelerate and move in any direction. The score ranges from zero to one, with higher scores indicating more balanced movement capabilities.
Most advanced robots today, including humanoid robots, four-legged robots, and drones, score below 0.6. Argus achieved a score of 0.91, bringing it close to the theoretical limit.
To develop the robot, the team used computer simulations to test more than 1,500 different robot designs. The simulations showed that robots with greater dynamic symmetry consistently performed better across a wide range of tasks. They were more stable, more energy-efficient, more resistant to damage, and better at handling difficult terrain.
The final design arranges the robot’s 20 legs around a geometric shape called a dodecahedron, which has 12 pentagonal faces. This arrangement helps distribute movement forces and camera coverage almost perfectly in every direction.
In real-world experiments on the Duke University campus, Argus quickly recovered after being pushed, carried a 10-pound payload while maintaining nearly full speed, and even climbed between two vertical walls by pushing against both sides with different groups of legs. It was also able to track and push large objects while continuously rolling.
The researchers believe Argus represents only the beginning of a new family of robots. Rather than copying animals or humans, future machines could be designed using mathematical principles that optimize movement and perception from the start.
According to the team, the goal is not simply to build robots that follow instructions. Instead, they hope to create machines that can explore, adapt, and help scientists discover new things about the world.
Argus serves as proof that this approach works. By abandoning traditional robot shapes and focusing on what a robot can do rather than what it looks like, researchers may have opened the door to a completely new generation of agile, resilient, and highly capable machines.
