Reprogrammable sequencing for physically intelligent underactuated robots

Area of Science:

• Robotics
• Mechanical Engineering
• Mechatronics

Background:

• Programming physical intelligence into mechanisms offers potential for resource-efficient machines.
• Underactuated mechanisms can perform complex tasks with fewer components.
• Autonomous motion adjustment in response to environmental interactions is a key challenge.

Purpose of the Study:

• To introduce a design approach for physically intelligent underactuated mechanisms.
• To demonstrate autonomous motion adjustment through passive reprogramming.
• To showcase devices that operate with minimal computational resources.

Main Methods:

• Harnessing multistability to sequence degrees of freedom.
• Utilizing mechanical stimuli from environmental interactions for passive reprogramming.
• Constructing a mass-sorting mechanism and a four-degree-of-freedom obstacle-avoiding robot.

Main Results:

• Demonstrated a mechanism that passively sorts objects by mass.
• Developed a robot capable of autonomous obstacle avoidance.
• Showcased devices operating without traditional computational architectures and using a single linear actuator.

Conclusions:

• Multistability provides a viable method for passive reprogramming of mechanism motion.
• Physically intelligent mechanisms can achieve autonomous behaviors with minimal resources.
• This approach offers a new paradigm for designing intelligent robotic systems.