Posted in | News | Consumer Robotics

Computationally Simple Robots Connect in Large Groups to Move Around and Transport Objects

Inspired from biological cells, a research team from Columbia University, MIT, and elsewhere has created computationally simple robots that are capable of connecting in huge groups to transport objects, move around, and complete other kinds of tasks.

Researchers have developed computationally simple robots, called particles, that cluster and form a single “particle robot” that moves around, transports objects, and completes other tasks. The work hails from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), Columbia University, and elsewhere. (Image credit: Felice Frankel)

Based on a project conducted by MIT, Cornell University, Columbia Engineering, and Harvard University researchers, the so-called “particle robotics” system contains a number of separate disc-shaped units, which are called “particles” by the researchers. Magnets around the perimeters of the particles connect them loosely, and each unit can perform only two things—contract and expand. (In the contracted state, each particle is roughly 6” and in the expanded state, it is roughly 9”). Meticulous timing of that motion causes the individual particles to pull and push each other in a coordinated motion. The cluster gravitates toward light sources through on-board sensors.

In a study recently reported in Nature paper, the team showed a virtual simulation of around 100,000 particles and a group of two dozen real robotic particles navigating through obstacles toward a light bulb. In addition, the researchers demonstrated that when objects are placed in the midst of a particle robot, it can transport that particular item.

These robots have the ability to form into various configurations and can move easily around obstacles and get in through constricted gaps. Most importantly, these particles do not interact directly with or depend on each other for their function, and therefore, particles can be subtracted or added without affecting the cluster. In their research paper, the investigators have demonstrated that even when a number of units malfunction, particle robotic systems still have the ability to complete the tasks.

The research paper points out a novel method to think about robots, which are typically developed for just one purpose—include various intricate components and cease working if any of these components break down. According to the researchers, robots made up of these basic parts could lead to more robust, flexible, and scalable systems.

We have small robot cells that are not so capable as individuals but can accomplish a lot as a group. The robot by itself is static, but when it connects with other robot particles, all of a sudden the robot collective can explore the world and control more complex actions. With these ‘universal cells,’ the robot particles can achieve different shapes, global transformation, global motion, global behavior, and, as we have shown in our experiments, follow gradients of light. This is very powerful.

Daniela Rus, Director, Computer Science and Artificial Intelligence Laboratory, MIT.

Rus is also the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science.

Other researchers who contributed to the paper are first author Shuguang Li, a CSAIL postdoc; co-first author Richa Batra, and corresponding author Hod Lipson, both from Columbia Engineering; Hyun-Dong Chang, David Brown, and Nikhil Ranganathan from Cornell University; and Chuck Hoberman from Harvard University.

For almost two decades, Rus has been working on modular, connected robots at MIT. These also include a cube robot that can expand and contract and can connect to others to move around. However, the group movement and configurations of the robots were restricted by the square shape.

In association with Lipson’s laboratory, where Li was a postdoc and later joined MIT in 2014, the research team went for disc-shaped mechanisms that can revolve around each other. In addition, these disc-shaped mechanisms can both connect and disconnect from one another, and form into various configurations.

A cylindrical base is provided on each unit of a particle robot, housing a tiny small motor, a battery, sensors that perceive light intensity, a communication component that relays and receives signals, and a microcontroller. A children’s toy known as a Hoberman Flight Ring is placed on top, and the inventor of this toy is one of the study’s co-authors. The Hoberman Flight Ring includes small panels joined in a circular formation which can be pushed back to contract and pulled to expand. Each panel is integrated with a couple of tiny magnets.

The key lies in programming the robotic particles to contract and expand in a precise sequence to pull and push the entire cluster toward a destination light source. To accomplish this feat, the team fitted each particle with an algorithm that examines the broadcasted data about the intensity of light from every other particle, without the necessity for direct communication between the particles.

Within a particle, the sensors spot the intensity of light from a light source; the intensity will be greater when the particle is closer to the light source. A signal, constantly broadcasted by each particle, shares its perceived level of intensity with all the other particles. For example, a particle robotic system determines the intensity of light on a scale of levels 1 to 10—that is, particles that are furthest from the light will register level 1, and those that are closest to the light will register level 10. The level of intensity consecutively corresponds to a particular time that should be expanded by the particle. Particles that experience the highest intensity—that is, level 10—will expand first. When those particles contract, the following particles in order, level 9, will expand subsequently. That timed motion of expansion and contraction occurs at every next level.

This creates a mechanical expansion-contraction wave, a coordinated pushing and dragging motion, that moves a big cluster toward or away from environmental stimuli.

Shuguang Li, Study First Author and Postdoc, Computer Science and Artificial Intelligence Laboratory, MIT.

Li added that the main component is the exact timing from a shared synchronized clock among the particles, facilitating efficient movement as much as possible: “If you mess up the synchronized clock, the system will work less efficiently.”

The team showed in videos how a particle robotic system containing real particles move and change its directions toward various light bulbs as they are switched on, and navigates via a gap between obstacles. The investigators have also described in their paper how simulated groups of around 10,000 particles preserve locomotion, at 50% of their speed, even when around 20% of units malfunctioned.

It’s a bit like the proverbial ‘gray goo,” said Lipson, a mechanical engineering professor at Columbia Engineering, referencing the science-fiction idea of a self-replicating robot that contains an infinite number of nanobots. “The key novelty here is that you have a new kind of robot that has no centralized control, no single point of failure, no fixed shape, and its components have no unique identity.”

Lipson further added that the next step would be to reduce the size of the components to create a robot comprising of countless numbers of very small particles.

Particle robots

(Video credit: MIT)

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.