CWI researcher Fredrik Jansson in Science Robotics on robot swarms – CWI Amsterdam

Growing biologically inspired shapes with hundreds of tiny robots

Scientists build self-organizing features into robot swarms to study shape formation

Hundreds of tiny robots can work together in a team to create bio-inspired shapes – without an underlying master plan based solely on local communication and movement. To achieve this, researchers from EMBL, CRG and Bristol Robotics Laboratory introduced the biological principles of self-organization into swarm robotics. Science Robotics published the results on December 19, 2018. Fredrik Jansson, who currently works at CWI, is one of the co-authors.

The robots used during the experiments. The shape of this particular crush is a hand made illustration of the technique. PHOTO: Reprinted with permission from AAAS

“We are showing that it is possible to apply nature’s self-organization concepts to human technologies such as robots,” says James Sharpe, group leader at EMBL Barcelona. “It’s fascinating because the technology is very brittle compared to the robustness that we see in biology. When a component in a car engine fails, it usually results in an inoperable car. On the other hand, if an element in a biological system fails, for example if a cell dies unexpectedly, this does not affect the entire system and is usually later replaced by another cell. If we could achieve the same self-organization and self-repair in technology, we could allow it to become much more useful than it is now. “Sharpe led the project initiated at the Center for Genomic Regulation (CRG) together with Sabine Hauert at the University of Bristol.


Form formation as seen in the robot swarms. Complete experiments took an average of three and a half hours. Inspired by biology, robots store morphogens: virtual molecules that carry the pattern information. The colors indicate the morphogen concentration of the individual robots: green indicates very high morphogen values, blue and purple indicate lower values ​​and no color indicates the virtual absence of morphogen in the robot. The morphogen concentration of each robot is transmitted to neighboring robots within a range of 10 centimeters. The overall pattern of spots that appear drives the displacement of robots to form protrusions that protrude from the swarm.
Video: Reprinted with permission from Slavkov, I., Zapata DC et al., Science Robotics (2018).

Turing’s rules

The only information the team installed in the coin-sized robots was basic rules for dealing with neighbors. In fact, they specially programmed the robots in the swarm so that they act similar to cells in a tissue. These “genetic” rules mimic the system responsible for the Turing patterns we see in nature, like the arrangement of fingers on a hand or the spots on a leopard. In this way, the project combines two of Alan Turing’s fascinations: computer science and pattern formation in biology.

The robots rely on infrared messages to communicate with neighbors within a 10 centimeter radius. As a result, the robots resemble biological cells, as they can only communicate directly with other cells that are physically close to them.

The swarm forms different shapes by moving robots from areas with low morphogen concentration to areas with high morphogen concentration – so-called “turing spots”, which leads to the growth of protrusions that protrude from the swarm. “It’s nice to see how the swarm grows in shapes. He looks pretty organic. It is fascinating that there is no master plan. These shapes are created through simple interactions between the robots. This differs from previous work where the shapes were often predefined. “Says Sabine Hauert.

Working with large swarms of robots

It is impossible to study swarm behavior with just a few robots. That’s why the team used at least three hundred in most of the experiments. Working with hundreds of tiny robots is a challenge in itself. They succeeded in doing this thanks to a special setup that makes it easy to start and stop experiments and to reprogram all robots with light at the same time. Over 20 experiments were performed with large swarms, each experiment lasting approximately three and a half hours.

As in biology, things often go wrong. Robots get stuck or run away from the swarm in the wrong direction. “That doesn’t happen in simulations, but only when you’re experimenting in real life,” says Ivica Slavkov, who shares the first authorship of the paper with Daniel Carrillo-Zapata.

All of these details made the project challenging. The early part of the project was done in computer simulations, and it took about three years for the real robot swarm to take on its first form. However, the limitations of the robots also forced the team to come up with clever, robust mechanisms to orchestrate the swarm patterns. Drawing inspiration from shape formation in biology, the team was able to show that their robot shapes can adapt to damage and repair themselves. The large-scale formation of the swarm is far more reliable than any of the small robots, the whole is greater than the sum of the parts.

Potential for real applications

While nature received inspiration for the swarm shape growth, the ultimate goal is to create large robotic swarms for real-world applications. Imagine hundreds or thousands of tiny robots growing shapes to explore a disaster environment after an earthquake or fire, or shaping into a dynamic 3D structure like a temporary bridge that automatically changes its size and shape to any building or building Can adapt terrain. “Since we were inspired by biological formation, which is known to be self-organized and robust against failure, such a swarm could still function if some robots were damaged,” says Daniel Carrillo-Zapata. However, there is still a long way to go before we see swarms like this outside of the laboratory.

James Sharpe (EMBL Barcelona) led the Swarm Organ Project, which was initiated at the Center for Genomic Regulation (CRG) when Sharpe was group leader there. Sabine Hauert (Bristol Robotics Laboratory and University of Bristol) was the key senior executive. Other employees were Fredrik Jansson (currently employed at Centrum Wiskunde & Informatica – CWI) and Jaap Kaandorp (University of Amsterdam – UvA).

The research leading to these results was funded by the European Union’s Seventh Framework Program (FP7) under Grant Agreement No. 601062 and the EPSRC Center for Doctoral Education in Future Autonomous and Robotic Systems (FARSCOPE) at the Bristol Robotics Laboratory.

More information

Article source
Slavkov, I., Zapata DC, et al. Morphogenesis in swarms of robots. Science Robotics, published online on December 19, 2018. DOI: 10.1126 / scirobotics.aau9178

This post was originally published on EMBL News.

Comments are closed.