Inspired by nature, new robot arm becomes more flexible in operation

Robotics Through plant-like oscillations, a plant-inspired robot controller was recently developed that can improve the performance of robotic arms in unstructured real-world environments, allowing the robot arm to reach a specified location or approach an object nearby.

The controller was developed by researchers from the Brain-Inspired Robotics (BRAIR) Laboratory, the BioRobotics Institute at the Istituto Superiore Sant'Anna in Italy, and the National University of Singapore, who drew inspiration from nature to artificially replicate biological processes, natural structures, or animal behaviors to achieve specific goals. This is because animals and plants are born with abilities that help them survive in their respective environments, and therefore can also improve the performance of robots outside of laboratory settings.

The controller, which uses information gained from proximity sensing embedded near the end effector to move toward a desired spatial target, was tested on a 9-DOF modular cable-driven continuous arm to reach multiple set points in space. The results are expected to enable deployment of these systems in unstructured environments.

According to Enrico Donato, one of the researchers who conducted the study, the soft robotic arm is a new generation of robotic manipulators that draws inspiration from the advanced manipulation capabilities demonstrated by "boneless" organisms such as octopus tentacles, elephant trunks, and plants, translating these principles into engineering solutions, thus forming a system composed of flexible and lightweight materials. These materials can undergo smooth elastic deformation to produce compliant and dexterous movements, properties that make the system physically robust and human-safe.

While soft robotic arms can be applied to a wide range of real-world problems, they are particularly useful for automation tasks that involve reaching desired locations that rigid robots might not be able to reach, and multiple research teams have recently been trying to develop controllers that would allow these flexible arms to handle these tasks efficiently.

Since most existing controllers for soft robotic arms were found to perform well primarily in laboratory settings, Donato and his colleagues set out to create a new type of controller that would also work in real-world settings, one that was inspired by the movement and behavior of plants.

"Contrary to the common misconception that plants don't move, plants actively and purposefully move from one point to another using growth-based movement strategies that are so effective that plants can colonize nearly every habitat on Earth, an ability that is lacking in the animal kingdom," Donato said. "Interestingly, unlike animals, plant movement strategies do not originate in the central nervous system, but rather take the form of complex decentralized computational mechanisms."

The team specifically used an artificial intelligence-based tool consisting of decentralized computational agents combined in a bottom-up structure. The novelty of the modified biomimetic controller lies in its simplicity. The researchers can use the basic mechanical functions of the soft robotic arm to generate the overall extension behavior. Specifically, the soft robotic arm consists of redundantly arranged soft modules, each of which is activated by a radially arranged triad of actuators. For this configuration, the system can generate six main bending directions.

So far, the researchers have tested their controller in a range of tests using a modular, cable-actuated lightweight soft robotic arm with nine degrees of freedom (9-DoF), and the results have been positive, with the arm being able to explore its surroundings and reach target locations more effectively than other control strategies have shown in the past.

In the future, the new controller could be applied to other soft robotic arms and tested in both lab and real-world environments to further evaluate its ability to cope with dynamic environmental changes. In the meantime, the R&D team plans to further develop their control strategy so that it can generate additional robotic arm motions and behaviors.