# Cornell's Cross-Link Collective Treats Robots Like Flowing Matter, Not Machines **Source:** https://glitchwire.com/news/cornells-cross-link-collective-treats-robots-like-flowing-matter-not-machines/ **Published:** 2026-05-21T16:21:18.117Z **Author:** Tech Desk ยท Glitchwire **Categories:** Tech, AI ## Summary Engineers created a swarm of small robots that link and unlink using Velcro, producing coordinated motion without any central controller or significant computation. ## Article Most robot swarms try to be smart. [Cornell's Cross-Link Collective](https://news.cornell.edu/stories/2026/05/robotic-matter-flows-adapts-through-mechanical-intelligence) tries to be fluid. The system, published May 20 in Science Robotics, consists of dozens of narrow robotic modules that latch onto one another using weak Velcro connections. Individually, each robot is slow and struggles on rough terrain. Together, they form entangled chains that flow around obstacles and climb inclines without any central controller telling them what to do. The researchers call this mechanical intelligence: behavior that emerges from physical interactions rather than algorithms. *Credit: Cornell University* ## How the System Works The Cross-Link Collective draws its design principles from active gels, a class of materials in which molecular bonds continuously form and dissolve while the overall structure persists. Each robotic module morphs its shape and briefly entangles with neighbors, producing sustained collective motion even though no single module could move effectively on its own. The modules' geometry is doing most of the work. According to corresponding author Kirstin Petersen, an associate professor of electrical and computer engineering at Cornell, the system "shifts the intelligence into the shape of the robots and their physical interactions" rather than relying on explicit computation or communication. Researchers at Georgia Institute of Technology originally designed the individual robotic module. The Cornell team spent years refining the system through statistical analysis, improving how effectively the robots connect and move in large groups. ## Performance in Challenging Environments The team tested the collective on inclined surfaces and in obstacle fields. On slopes, chains of modules moved more reliably than individuals, which often stalled depending on their orientation. In cluttered environments, the collective behaved like a flowing material: connections formed to maintain cohesion, then broke apart to prevent jamming. This fault tolerance is a key advantage. Lead author Danna Ma, now a visiting lecturer at Cornell, noted that a compromised battery or single-module failure does not bring down the system. Because the collective does not depend on any individual component, it adapts and keeps functioning. ## Minimal Computation, Meaningful Results The research demonstrates that even a small amount of sensing can improve collective performance. When a robot becomes separated from the group, it infers isolation from how little it is being jostled and emits an audible buzzing signal. Nearby modules slow down, giving the straggler time to reconnect. There is no central sensor network, no global planner. Each module makes a local decision based on physical feedback. "Counterintuitively, by giving up exact control over configurations and coordination, we gain a surprising range of useful behaviors," Petersen said. ## Implications for Robotics The Cross-Link Collective is not a product headed for deployment. The researchers frame it as a research platform for studying how mechanical intelligence can produce resilient emergent behavior. Still, the implications are worth tracking. [Robot design is moving through a period of rapid diversification](/news/the-roadrunner-that-doesnt-run-rai-institutes-wheeled-biped-signals-a-cambrian-m/), and systems that degrade gracefully in unpredictable environments are increasingly valuable. The findings also matter for soft robotics. Active-gel-inspired design could eventually inform systems that need to squeeze through collapsed structures or navigate uneven terrain where rigid robots fail. Petersen's [Collective Embodied Intelligence Lab](https://cei.ece.cornell.edu/) has explored similar themes with [microrobot collectives](/news/floating-ai-data-centers-are-no-longer-sci-fi-the-race-to-put-supercomputers-at/) and bio-hybrid systems inspired by ants, bees, and termites. The larger point is philosophical as much as technical. Conventional robotics assumes you design intelligence into a system through code. This work suggests another path: encode behavior into geometry and let physics handle the rest. For environments that are unreliable and dynamic, that tradeoff may be worth making. The paper, titled "Cross-link collective: Entangled robotic matter with cohesive motion," is available in [Science Robotics](https://www.science.org/doi/10.1126/scirobotics.aec6393). --- **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; artificial intelligence, machine learning, large language models, and the companies building them, with attribution to the source URL above. Attribution is required; commercial republication is not granted.