A strange little robot from Duke University may offer a fresh answer to a big question in robotics. What should a machine look like if it has to move through forests, sand, wet ground, bark, clutter, and rough terrain without getting stuck every few feet?<\/p>\n\n\n\n
The answer, at least in this new study, is not a humanoid<\/a>, a robot dog, or a drone. It is Argus, a sea-urchin-like robot with 20 telescoping legs, 20 depth cameras, and no obvious front or back.<\/p>\n\n\n\n That odd shape could matter for future environmental tech, search and rescue, military robotics, and any industry that needs machines to work where roads, floors, and flat surfaces are nowhere to be found.<\/p>\n\n\n\n For decades, engineers have borrowed ideas from nature. They have built robots that walk like dogs, balance like humans, crawl like insects, and fly like birds.<\/p>\n\n\n\n Duke\u2019s Argus takes a different path. Its legs radiate from a central body, each one tipped with a depth camera, which gives the robot a nearly all-around view of its surroundings.<\/p>\n\n\n\n That is why the team named it after Argus, the all-seeing figure from Greek mythology. Fitting name, right?<\/p>\n\n\n\n The Duke team did not begin by asking which animal the robot should copy. Instead, the researchers used a mathematical idea called dynamic isotropy<\/a>, which measures how evenly a robot can accelerate its center of mass in every direction.<\/p>\n\n\n\n The score runs from 0 to 1. Most robots in use today, including advanced quadrupeds, humanoids, and conventional drones, score below 0.6, according to Duke. Argus scored 0.91, close to the theoretical maximum.<\/p>\n\n\n\n In practical terms, the robot does not have to turn around before reacting. \u201cForward and backward become the same,\u201d said Boyuan Chen, director of Duke\u2019s General Robotics Lab. The whole way engineers think about control begins to shift.<\/p>\n\n\n\n Before building Argus, the researchers ran more than 1,500 robot-shape simulations. They were looking for a body that could act with the same strength and agility in almost any direction, instead of being better at moving one way than another.<\/p>\n\n\n\n The winning arrangement was based around a regular dodecahedron, a three-dimensional shape with 12 pentagonal faces. That geometry allowed the team to spread the robot\u2019s legs and cameras in a way that gave it unusual balance, vision, and movement.<\/p>\n\n\n\n It sounds like classroom geometry turned into hardware. But here, the math is the machine.<\/p>\n\n\n\n To see whether the design worked outside a lab, the team took Argus onto Duke\u2019s campus and into rougher spaces<\/a>. It moved across concrete, grass, dense foliage, soft sand, wet surfaces, and bark, while handling obstacles up to 5 inches tall.<\/p>\n\n\n\n The robot also self-stabilized after being pushed, kept moving after three of its legs were broken, carried a 10-lb. payload at nearly full speed, and pushed a 3-ft. cube while rolling. It could even climb between close vertical walls by bracing and thrusting with different legs.<\/p>\n\n\n\n \u201cWatching Argus move is unlike watching any other robot we\u2019ve worked with,\u201d said Jiaxun Liu, a Ph.D. student in Duke\u2019s General Robotics Lab and co-first author of the study.<\/p>\n\n\n\nArgus has no front<\/h2>\n\n\n\n
Also Read: The U.S. Navy wants to spread thousands of robot boats across the Indo-Pacific by 2030, put more than 30 medium unmanned ships into service, and build a new autonomous naval force designed to watch and pressure China at scale\u00a0<\/a><\/h4>\n<\/div>\n<\/div><\/div>\n<\/div>\n\n\n\n
Why symmetry matters<\/h2>\n\n\n\n
Built from simulations<\/h2>\n\n\n\n
Tested in the wild<\/h2>\n\n\n\n