MIT researchers have developed "FloatForm," a swarm of autonomous, square robotic boats designed to assemble themselves into reconfigurable structures on water. Each unit measures 21 centimeters square and functions as an independent agent equipped with its own thrusters, sensors, and magnetic latches. The system allows for the creation of temporary floating infrastructure, such as bridges, stages, or platforms, which can form, disassemble, and move to new configurations with minimal human intervention. Unlike traditional self-assembling robot systems that rely on a centralized computer for every movement, the FloatForm approach utilizes a decentralized coordination algorithm. While a central planner assigns final positions to ensure geometric precision, the individual robots manage navigation, collision avoidance, and environmental adaptation by exchanging positional data with their immediate neighbors. This parallelism enables the swarm to remain highly scalable, as computational complexity depends only on local neighbor interactions rather than the total size of the group.

The project draws inspiration from fire ants, which form living, leaderless rafts to survive flooding. To achieve stable connections, the researchers developed a latching mechanism using an origami-inspired auxetic structure (geometry that contracts uniformly in all directions). A servo motor pulls magnets inward to release or pushes them outward to grab neighboring units across 10 to 15-centimeter gaps. Once latched, the mechanism holds its state without consuming additional power, which preserves battery life for computation and movement. Testing in an indoor tank demonstrated that the robots could reliably gather from random positions, latch into rigid structures, and coordinate as a single vessel during collective transport, with each mission taking four to eight minutes to complete. In 10 trials, the system completed missions without human intervention 90 percent of the time with four robots and 70 percent with eight.

The research, published in Nature Communications on July 9, 2026, builds upon the earlier Roboat project conducted by MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Senseable City Lab. The team noted that scaling the technology for real-world canals or harbors will require transitioning from ultrasonic indoor positioning to GPS or vision-based sensing and potentially reinforcing latches to withstand natural water disturbances. Beyond urban applications like floating markets or adaptive infrastructure, the researchers suggest the technology could support offshore inspection, environmental monitoring, and emergency response operations. The coordination framework was designed to be sensor-agnostic, allowing the system to adapt to different operational environments and sensor suites in the future.

MIT researchers have developed "FloatForm," a swarm of autonomous, square robotic boats designed to assemble themselves into reconfigurable structures on water. Each unit measures 21 centimeters square and functions as an independent agent equipped with its own thrusters, sensors, and magnetic latches. The system allows for the creation of temporary floating infrastructure, such as bridges, stages, or platforms, which can form, disassemble, and move to new configurations with minimal human intervention. Unlike traditional self-assembling robot systems that rely on a centralized computer for every movement, the FloatForm approach utilizes a decentralized coordination algorithm. While a central planner assigns final positions to ensure geometric precision, the individual robots manage navigation, collision avoidance, and environmental adaptation by exchanging positional data with their immediate neighbors. This parallelism enables the swarm to remain highly scalable, as computational complexity depends only on local neighbor interactions rather than the total size of the group.

The project draws inspiration from fire ants, which form living, leaderless rafts to survive flooding. To achieve stable connections, the researchers developed a latching mechanism using an origami-inspired auxetic structure (geometry that contracts uniformly in all directions). A servo motor pulls magnets inward to release or pushes them outward to grab neighboring units across 10 to 15-centimeter gaps. Once latched, the mechanism holds its state without consuming additional power, which preserves battery life for computation and movement. Testing in an indoor tank demonstrated that the robots could reliably gather from random positions, latch into rigid structures, and coordinate as a single vessel during collective transport, with each mission taking four to eight minutes to complete. In 10 trials, the system completed missions without human intervention 90 percent of the time with four robots and 70 percent with eight.

The research, published in Nature Communications on July 9, 2026, builds upon the earlier Roboat project conducted by MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Senseable City Lab. The team noted that scaling the technology for real-world canals or harbors will require transitioning from ultrasonic indoor positioning to GPS or vision-based sensing and potentially reinforcing latches to withstand natural water disturbances. Beyond urban applications like floating markets or adaptive infrastructure, the researchers suggest the technology could support offshore inspection, environmental monitoring, and emergency response operations. The coordination framework was designed to be sensor-agnostic, allowing the system to adapt to different operational environments and sensor suites in the future.