Teaching robots to play as a team with nature

Intelligent Computing (2022). DOI: 10.34133 / 2022/9761694 “width =” 800 “height =” 383 “/>

Results for the unimodal scenario. Illustration of the design method and comparison with multi-agent simulations for the unimodal scenario: (a) describes the stationary distribution and (b) the expected change. Credit: Intelligent computing (2022). DOI: 10.34133 / 2022/9761694

Algae flourish, birds flock and insects swarm. This mass behavior by individual organisms can provide a separate and collective good, such as improving the chances of good mating propagation or providing safety. Now, researchers have harnessed the self-organizing capabilities needed to reap the benefits of natural swarms for robotic applications in artificial intelligence, computing, search and rescue, and more.

They released their method in August. 3 inches Intelligent computing.

“Designing a set of rules that, once executed by a swarm of robots, results in a specific desired behavior is particularly challenging”, said corresponding author Marco Dorigo, professor in the artificial intelligence laboratory, called IRIDIA, of the Universit√© Libre de Bruxelles, Belgium. “Swarm behavior is not a one-to-one map with simple rules performed by individual robots, but rather results from the complex interactions of many robots running the same set of rules.”

In other words, robots must work together to achieve the sum of discrete contributions goal. The question, according to Dorigo and his co-authors, Dr. Valentino and prof. Hamann, is that conventional design for individual units to achieve a collective goal is bottom-up, requiring refinements by trial and error that can be costly.

“To address this challenge, we propose a new global-local design approach,” said Dorigo. “Our key idea is to compose a heterogeneous swarm using different groups of behavioral agents in such a way that the resulting swarm behavior approaches a user input that represents the behavior of the entire swarm.”

This composition involves the selection of individual agents with predetermined behaviors that researchers know will work together to achieve the target collective behavior. They lose the ability to program individual units locally, but according to Valentini, Hamann and Dorigo it is worth it. They pointed to the example of a surveillance activity, where a swarm may need to monitor a facility that requires more internal monitoring during the day and more external monitoring at night.

“The user provides a description of the desired swarm allocations as a spatial probability distribution of all possible swarm allocations: more agents inside during the day, more outside at night or vice versa,” said Valentini.

The user would define the target behavior by changing the number and location of the distribution modes, with each mode corresponding to a specific allocation, such as 80% of agents inside, 20% outside during the day and the 30% indoors, 70% outdoors at night. This allows the swarm to change behavior periodically and autonomously, predetermined by the modalities set, as circumstances change.

“While it is difficult to find the exact control rules for robots to make the swarm behave the way we want, it is possible to achieve desired swarm behavior by combining different sets of control rules that we already understand,” said Dorigo. “Swarm behaviors can be macroscopically designed by mixing robots with different predefined rule sets.”

This is not the first time that Dorigo has turned to nature to improve IT approaches. Previously he developed the ant colony optimization algorithm, based on how ants navigate between their colonies and food sources, to solve difficult computational problems that involve finding a good approximation of an optimal path on a graph.

Although Dorigo first proposed this approach for a relatively simple problem, it has since evolved as a means of addressing a variety of problems. Dorigo said he intends to take the swarm methodology in a similar direction.

“Our next immediate step is to demonstrate the validity of our methodology through a broader set of swarm behaviors and move beyond asset allocation,” said Dorigo. “Our ultimate goal is to understand what makes this possible by formalizing a generic theory to allow researchers and engineers to design swarm behaviors without going through the meticulous process of trial and error.”

Less communication between robots allows them to make better decisions

More information:
Gabriele Valentini et al, Global-to-Local Design for Self-Organized Task Allocation in Swarms, Intelligent computing (2022). DOI: 10.34133 / 2022/9761694

Provided by Intelligent Computing

Citation: Teaching Robots to Team Up with Nature (2022, September 21) Retrieved September 21, 2022 from https://techxplore.com/news/2022-09-robots-team-players-nature.html

This document is subject to copyright. Outside of any commercial fairness for private study or research purposes, no part may be reproduced without written permission. The content is provided for informational purposes only.

Leave a Comment

%d bloggers like this: