Clever swarms of bacteria inspire robot builders

A single bacterium is not very smart, but a colony of bacteria behaves quite advanced. Researchers have now developed a new model that can also be applied in swarms of robots.

In this computer simulation, a swarm of bacteria finds its way through a jungle of obstacles

Intelligent swarm
The researchers at Tel Aviv University in Israel have developed a calculation model that better explains how bacteria move in a swarm and can also be applied in computers, artificial intelligence and robots [1]. The principles that bacteria use to move in a swarm can also be used very well to make computers, artificial intelligence and robots exhibit swarming behavior. Applications that the researchers mention are the development of medical nanorobots that deliver drugs into the body or collect information about consumers on the internet.

Why do stupid bacteria survive in a complicated environment?
Although bacteria are small and simple, they have superior survival skills. The way they make decisions and their collective behavior allows them to thrive and even spread in harsh living conditions. As extremely simple organisms, bacteria are of course incapable of high-quality information-processing processes. In a complex environment, such as inside people or animals, bacteria would get confused. At least, it was assumed.

'Self-confidence' saves bacteria
This assumption turns out to be wrong. In a startling discovery, the researchers determined that bacteria do, in fact, have superior survival skills. They find food much earlier and avoid danger much faster than more complex cooperating organisms such as amoebas or fish. Their secret: a lot of 'self-confidence'.

Many swarms of more complex organisms are often plagued by false positive feedback, which is common in complex terrain. This happens when a subgroup of the swarm, based on wrong information, sends the whole group in the wrong direction. Bacteria do not make this mistake because they communicate via molecular, chemical or mechanical means.

Only 'ask for advice' when things go wrong
Depending on how much 'self-confidence' they have, bacteria can adapt their interaction with their fellow bacteria. When an individual bacterium finds a useful path, it pays less attention to the signals of others. If the cell encounters a less favorable path, it will 'ask for advice' from other cells. Each cell follows the same strategy, causing the group as a whole to find its way through extremely complex terrain through trial and error. Exactly this behavior, which requires little capacity and short-term memory, was incorporated into the researchers' model. This principle can also lead to new, more efficient technology.

Robots borrow bacteria strategy
Robots also often have to navigate in complex environments, such as just about any conceivable environment outside of a lab. They also have to communicate with each other. At the moment this is calculated with difficulty, which of course requires a lot of computing resources from computers. Unraveling the secrets of swarms of bacteria could help us develop a new generation of robots that can move in an environment without too much built-in cleverness and adapt to changing conditions, the researchers say [2].

1. Adi Shklarsh et al., Smart Swarms of Bacteria-Inspired Agents with Performance Adaptable Interactions, PLoS Computational Biology, 2011 (free access)
2. Smart Swarms of Bacteria Inspire Robotics Researchers, AFTAU, 2011

Leave a Comment