autoevolution
 

Team Planning Algorithm Helps Robots Succeed in Rescue or Surveillance Missions

You can try to create a team with only the best of the best, but, if there’s no coordination, the results are not going to be good. Efficient team work is what it’s all about. And the same goes for robots.
Researchers at MIT have developed a team planning method that will improve the way robots and drones work together 1 photo
Photo: Massachusetts Institute of Technology
It starts off with one cute robot that’s delivering your packages and, soon enough, we won’t be surprised to see teams of robots and drones performing various tasks that help the community. Of course, scientists are already a few steps ahead of that, working on how to improve communication within robot teams and make sure that they get the best results in the most effective way.

This is what a team (it figures) of scientists at the Massachusetts Institute of Technology (MIT) wants to develop. Until now, when it came to robot teams, the focus on quantity and gathering as much information as possible. But that comes at a higher cost in terms of energy consumption.

This new algorithm, called Distributed Local Search, is a way of optimizing data collection while using less energy. Typically, robots within a team would each plan their trajectory and then go for it, one by one, which takes more time and energy. With this new planning method, each robot still gets to generate a possible trajectory, only this time they have to share it with the rest of the team. Then, the algorithm accepts or rejects each trajectory, based on the overall “objective function”. This means analyzing whether gathering that particular set of data would be worth the energy cost or not.

By using this algorithm, robots no longer have to waste precious energy just to get more information that won’t make a difference. According to the researchers, who have tested this method on a simulated team of 10 robots, Distributed Local Search takes a longer running time, but it saves energy and helps robots accomplish their mission.

This can have a positive impact on real-world missions such as aerial surveillance, ocean monitoring or search-and-rescue, where limited resources need to be used wisely. In order to get there, the next step is testing this algorithm on teams of robots and drones.
If you liked the article, please follow us:  Google News icon Google News Youtube Instagram X (Twitter)
About the author: Otilia Drăgan
Otilia Drăgan profile photo

Otilia believes that if it’s eco, green, or groundbreaking, people should know about it (especially if it's got wheels or wings). Working in online media for over five years, she's gained a deeper perspective on how people everywhere can inspire each other.
Full profile

 

Would you like AUTOEVOLUTION to send you notifications?

You will only receive our top stories