Flocks of robots work better with less opinion

According to one study, when decisions within a group are made on a small number of machines, a swarm of robots can effectively tackle forest fires and other hazards.

LThis study, edited by Dr. Andreagiovanni Reina of the University of Sheffield’s School of Computer Science, Science robotics You can improve how swarms of robots work together, adapt to changing environments, and make advanced decisions faster.

This study is said to disprove the widely accepted theory that more connections between robots lead to more effective information exchange.

Click here for more information on robotics

A team of researchers from UCL and IRIDIA at the Free University of Brussels, Belgium, by studying how a swarm of small robots moved around and reached an agreement on the best areas to gather and explore. I found a discovery.

According to the survey, each robot evaluated the environment individually, made its own decisions about the optimal area, and broadcasted its opinions to the rest of the herd. Then all the robots in the herd periodically select a random rating broadcast by another robot in the herd and use it to be the best known protocol called the voter model. Updated opinions on the area of. Once all the robots went through this process, the herd reached an agreement on the best areas to collect and explore based on the opinions of each robot.

By using this protocol, the team found that swarms of robots were slow to adapt to changes in the environment when better sites emerged.

Increasing connections between robots does not necessarily lead to effective information exchange (Image: University of Sheffield)

Researchers then found that when the robot communicated with other robots within a 10 cm range rather than broadcasting a message throughout the group, the herd adapted much more quickly to changes in the environment and was optimally available. I found that I could select a region.

In a statement, Dr. Reina said: “A swarm of robots has great potential to help access places that are too dangerous or inaccessible to humans. For example, flying over forest fires that are too wide or dangerous for humans to tackle alone. You can monitor how the fire spreads and determine where you need the most help.

“But what if the fire suddenly turns and needs urgent help elsewhere? A swarm of robots quickly adapts to this change and finds out where urgent help is needed. We need to be able to identify. This is what our research is helping. Our findings show that more responsive robots can make the right decisions much faster than they are today. It can be used to develop a herd. “ Flocks of robots work better with less opinion

Back to top button