Breathing life-like motion into a self-propelling robot
July 31, 2024 | Bharti Dharapuram
A recent study programmed a palm-sized robot to perform motions ranging from passive (Brownian) to active dynamics (Active Brownian and Run-and-Tumble). Shown in the image are the tracks of the robot reconstructed from overhead video footage, which was analyzed by the researchers.
Many bacteria use their energy to perform complex behaviours that allow them to move toward food while avoiding harmful repellents. The dynamics of such active systems are governed by rules in physics that can be adapted for human applications. A recent collaborative study tuned a palm-sized robot to perform a range of motions that can allow us to study active dynamics in the lab.
The research group behind the study includes Somnath Paramanick and Nitin Kumar from the Indian Institute of Technology (IIT) Bombay, Arnab Pal from the Institute of Mathematical Sciences, Chennai, and Harsh Soni from IIT Mandi.
Researchers programmed the robot to control the velocities of its two independently moving wheels and mounted it with sensors to detect light. Based on these signals, the robot can exhibit various dynamics, including those characteristic of living systems. The researchers showed that this robot switches between different kinds of motions, avoids obstacles and navigates towards a target by sensing a gradient in light.
Active entities consume energy to perform work and are ubiquitous in nature. These can be bacteria using flagella to propel themselves, a protein moving through a cell, or a bird flying in the sky. Scientists are interested in the dynamics of these self-propelling particles to understand emergent phenomena such as complex collective behaviour or self-evolved optimal navigational strategies. However, it is difficult to manipulate living systems in the lab, which can be a limiting step in understanding the underlying principles of such behaviour.
Therefore, when Kumar approached Pal with affordable and easily maneuverable robots, it presented many possibilities. Pal’s research interests are in understanding search strategies across various scales. These range from repair proteins looking for DNA mutations to fix, homing pigeons traveling back to their nest, or a computer algorithm scanning a tree of possibilities in search of a solution. Pal and his colleagues were motivated to build a robust in-house system to study such behaviour using Kumar’s robots.
“First, we wanted to see if it (the robot) actually performs different types of motion,” Pal says. The researchers did this by programming the robot so that signals reaching its wheels could elicit different kinds of motion. “By adding stochasticity to the signals, we have been able to cover many different classes of stochastic processes,” he says. They replicated Brownian motion seen in diffusion of gas molecules, run and tumble motion used by bacteria to align and move in a medium, and an active form of Brownian motion observed in the motion of Janus particles. After demonstrating a variety of passive and active dynamics, the researchers programmed the robots to switch between these movements using environmental cues. The robot efficiently navigated to a target by avoiding obstacles using infrared sensors and orienting itself based on differences in light intensity.
The next step would be to see whether this robust lab based system can offer valuable insights into the physics of homing – the incredible ability to navigate back to a home or a base from unfamiliar places, a behavior which is surprisingly ubiquitous across the animal kingdom, Pal says.
