Direct contact of random objects from the open environment to the panel surface of an electronic device may reduce the work efficiency and cause permanent damage. However, there is a possible way to solve this problem, notably by implementing an adaptive structure design inspired by plants. TheMimosa pudicaplant provides several interesting information on its adaptability. Various studies have been conducted on the electrical properties of its organs explaining the phytoactuator and phytosensor cells that function within it. We combined the use of sensors, actuators, and synthetic excitable tissue as the first robot model purposed to mimic the behavior of theM. pudicaplant. The Computer vision method was used to measure leaf angular movement and collected it as plant behavior data based on the mechanical stimulus experiment. The Robot structure has eight arms equipped with sensors, servo motors, and microcontrollers that are operated with two activation system models approach. The first model could imitate the stimulus process received by electronic circuits that generate action potential signals with a maximum voltage of 4.71-5.02 V and a minimum voltage of -5.33 to -3.45 V that propagated from node to node. The second model involves a trained artificial neural network model with a supervised learning pattern that provides 100% accuracy when choosing movement output based on the given combination. This robot imitates theM. pudica's intelligent sensing capabilities and its ability to change the structure shape based on the thygmonasty experiments data which could provide an overview of how plants process information and perform hazard avoidance actions efficiently. Future applications for the technology inspired by the plant's self-defense mechanisms are adaptive intelligent structures that can protect against harmful conditions, particle contamination, and adjusting panel structure to search for desired environmental parameters.
Keywords: Mimosa pudica; action potential; biomimetic robot; neural network; plant-inspired robot.
© 2022 IOP Publishing Ltd.