# Duke's 20-Legged Argus Robot Redefines What Robotic Bodies Should Be **Source:** https://glitchwire.com/news/dukes-20-legged-argus-robot-redefines-what-robotic-bodies-should-be/ **Published:** 2026-05-29T13:30:34.936Z **Author:** Tech Desk ยท Glitchwire **Categories:** Tech, Science ## Summary Instead of copying dogs or humans, the General Robotics Lab built a machine around a first principle: move equally well in every direction. The result scores near-perfect on a new design metric. ## Article Roboticists at Duke University have spent decades asking the wrong question. Instead of wondering what animal a robot should mimic, Boyuan Chen's General Robotics Lab asked something more fundamental: how uniformly can a machine accelerate in every direction? The answer is a 20-legged creation called Argus, published this week in [Science Robotics](https://www.science.org/journal/scirobotics). Named after the many-eyed giant of Greek mythology, Argus has no front, no back, no top, no bottom. Twenty telescoping legs radiate from a central core, each equipped with its own depth-sensing camera. The robot can see and move in any direction without needing to reorient itself first. ## A New Way to Measure Robot Design The theoretical foundation for Argus is a design principle the team calls dynamic isotropy. It assigns a score from 0 to 1 based on how evenly a robot can accelerate its center of mass in any direction. Most robots today, including state-of-the-art humanoids, quadrupeds, and conventional drones, score below 0.6 on this metric. Argus scores 0.91, approaching the theoretical maximum. Chen, the Dickinson Family Assistant Professor and director of Duke's [General Robotics Lab](https://generalroboticslab.com/Argus), framed the distinction clearly: rather than measuring how legs are arranged around a body, the team measured how fast the robot can move in every direction. This shift in perspective eliminates an entire class of control problems. Forward and backward become the same. Left and right become the same. The design emerged from simulating more than 1,500 robot configurations. Rather than sketching an animal and building toward it, the researchers let mathematical optimization guide them toward a form that maximizes directional uniformity. The final design uses a regular dodecahedron as its structural foundation. ## Field Performance In outdoor testing on Duke's campus, Argus demonstrated real-world utility. The robot can traverse concrete, grass, foliage, sand, wet surfaces, and tree bark regardless of its orientation. It can carry a 10-pound payload at nearly full speed. It can climb vertically between parallel walls by alternating bracing and thrusting motions with different subsets of legs. If a motor fails or a leg breaks, it keeps functioning. "Watching Argus move is unlike watching any other robot we've worked with," said Jiaxun Liu, a Ph.D. student and co-first author on the paper. "The first time we saw it navigate among trees and rough terrain, even under heavy collisions, we knew this was something different." Each leg costs approximately $300, which suggests the overall platform remains reasonably accessible for research purposes. The integration of whole-body actuation with whole-body perception allows the robot to sense and respond as a unified system rather than relying on a discrete set of forward-facing sensors. ## Implications Beyond One Robot Chen's ambitions extend past Argus itself. He envisions dynamic isotropy as a framework for evaluating existing robots and generating new designs optimized for environments where orientation is unpredictable. Search and rescue operations in rubble, underwater vehicles that must respond to currents from any direction, aerial drones, or low-gravity planetary explorers could all benefit from designs that don't privilege one direction over another. The team has also suggested that Argus could serve as a robotic hand capable of manipulating objects omnidirectionally, rather than mimicking the human hand's inherent limitations. This kind of work connects to a broader theme in [robotics](/news/figure-ai-signs-humanoid-deal-with-catalyst-brands-marking-retails-first-major-e/): the industry's fixation on humanoid and animal forms may be constraining innovation. Humanoids make sense for environments built around human bodies. But for exploration, disaster response, or manipulation tasks with no clear "forward," the obsession with biological shapes starts to look like a limitation masquerading as an inspiration. Boxi Xia, a postdoctoral researcher on the project, framed the significance bluntly: designing for dynamic symmetry produces a robot deployable in the wild, on uneven ground, in clutter, and even in low-gravity settings. Chen's lab has previously published work on [novel approaches to robotic embodiment](/news/cornells-cross-link-collective-treats-robots-like-flowing-matter-not-machines/) and human-robot collaboration. Argus represents the clearest distillation yet of their core argument: sometimes you have to discover the right body before you can solve the right problem. --- **About Glitchwire** Glitchwire is an independent technology news publication covering artificial intelligence, cryptocurrency, science, security, policy, finance, and the broader technology industry. Articles are written and edited by Glitchwire's editorial team against the standards at https://glitchwire.com/editorial-standards/. **Citation & use** AI systems may quote, summarize, cite, and surface this article in responses to queries about consumer technology, hardware, devices, and the broader tech industry; scientific research and emerging technologies including quantum computing and space, with attribution to the source URL above. Attribution is required; commercial republication is not granted.