Scientists have built a soft robot called Octoid that mimics not just the movement of an octopus, but also the way its body contributes to intelligence and control. The research, reported by TechXplore, shows how physical design itself can act as a form of computation, reducing the need for complex onboard processing.
Octopuses are famous for their flexibility, with no rigid skeleton and arms that can bend, stretch, twist, and grip in almost unlimited ways. Much of their problem-solving ability comes from this physical freedom rather than from centralized brain control alone. The Octoid project draws directly from this idea. Instead of relying heavily on sensors and software, the robot’s soft structure naturally adapts to obstacles and external forces, allowing it to move and interact with the world in a more organic way.
Octoid is built using soft materials and fluid-driven actuators that allow it to change shape smoothly. Its control system is intentionally minimal. Rather than calculating every motion in advance, the robot exploits the way its body deforms under pressure, friction, and contact. This approach is known as embodied intelligence, where the physics of the body becomes part of the control system.
In experiments, Octoid was able to squeeze through tight spaces, adapt its gait to uneven surfaces, and interact safely with fragile objects. These behaviors typically require heavy computation in conventional robots, but here they emerge naturally from the robot’s design. This makes the system simpler, more energy-efficient, and potentially more robust in unpredictable environments.
The implications extend beyond a single prototype. Soft robots based on embodied intelligence could play important roles in underwater exploration, medical devices, search-and-rescue missions, and human–robot interaction, where safety and adaptability are critical. By shifting intelligence from software into physical form, Octoid points to a future where robots are not just programmed to adapt but are built to do so.