Stanford Seminar - Engineering physical principles of embryonic morphogenesis in robotic collectives

Morphogenesis in Robots

• Morphogenesis is the process by which an embryo changes shape by rearranging its cells.
• Current collective robots can change shape but are limited to perimeter motions.
• The goal is to create a robot that can rearrange its cells in a tight-packed situation throughout the material.
• The basic way to rearrange the topology of cells is called the T1 transition or intercalation.

Design Challenges

• Creating surface flows similar to cells.
• Adhesion between cells.
• Polarizing the robots so they know which direction to elongate.

Robot Design

• The final design of the robots included a custom circuit board, cheap motors, batteries, gears, and sensors.
• To create surface flows, the team used gears driven by motors along the surface of the robot.
• For adhesion, magnets were embedded in the robots and allowed to rotate to match the orientation of neighboring robots.
• To polarize the robots, polarized light was used. Each robot had photodiodes with polarized film on top, and a second polarized film was placed above the arena.

Experiments and Observations

• The robots were able to successfully perform T1 transitions when the polarized light was shone on them.
• The researchers conducted experiments to understand how the shape and strength of a collective of robots change in response to different parameters.
• They observed a sharp threshold in the force required for motion, both when varying the average force and the amplitude of fluctuations.
• The researchers found that as the amplitude of fluctuations increases, the average force required for motion decreases.
• A steep bimodal distribution of force pulses is most effective for shape change.

Robot Capabilities

• The robots can be turned on and off, and their stiffness and strength can be controlled by varying the inter-unit force and vibrations.
• The power consumption of the robots is reduced when they are fluctuating, but the energy required is similar.
• The robots can be controlled collectively, and they can rearrange themselves and change shape in response to light stimuli.
• The collective of robots can be turned into a viscoelastic material that can be deformed and reshaped.
• The robots can exhibit spatial control, where one side of the collective can be fluctuating while the other side is not.

Potential Applications

• Soft robotics
• Self-healing materials
• Programmable matter

Challenges and Future Research

• Scaling down the size of the robots
• Enabling 3D motion
• Incorporating distributed computation and communication capabilities