Flick on the light when a family of cockroaches are scuttling across your kitchen floor and they’ll swiftly disperse – only to regroup in the walls.

To understand how cockroaches – and other animals – work as a group, researchers are sending in robots. These can help reveal group dynamics. Bee bots have been used to study waggle dances in honeybee hives, for example.

Others are looking at how robots can infiltrate colonies and influence behaviour. Bee bots could protect pollinating bees against mites – or keep a hive updated with weather forecasts, releasing pheromones to prevent a new brood hatching when bad weather is likely to prevent foraging. Robots can even be used to encourage groups of animals to do things they normally would not: a robo-roach can trick cockroaches into venturing into the light, for example.

But getting the robots to behave in a way that makes them blend in can be tricky. In previous work, JosĂ© Halloy at Paris Diderot University in France and his colleagues programmed their robot cockroaches largely by hand. But this is hard – and cannot easily be adapted for use with other types of animal.

So Halloy’s team has now developed a way to generate the robots’ behaviour automatically using a mix of descriptions of cockroach habits, combining models of individuals’ movement with group activity. They then used evolutionary algorithms to optimise the models.

Behaviour is what counts

Programming the robots to behave like individual insects is not the best way, says team member Nicolas Bredeche at Paris-Sorbonne University. “You don’t know if these small details will capture the global behaviour of the cockroaches when they are together,” he says.

The team tested their generated behaviours in a computer simulation in which a mixed group of 45 cockroaches and five robots had to cooperate to make a collective choice between two shelters. Generating insect-mimicking behaviour automatically was a lot quicker than doing it by hand – and led to more lifelike behaviour. They found that the mixed group acted like a real cockroach group – grouping along walls in realistic ways, for example.

The team think that the approach could be used to generate behaviours for mimics of other social species, such as honeybees, fruit flies, birds and fish. Behaviour is much more important than looks, says Bredeche. A fish will accept a robot as another fish if it acts like one, even if it doesn’t look the part. And group level behaviour matters even more than individual actions. The movement and paths followed by the robo-roaches did not need to match those of real cockroaches for them to fit in.

Using robots that mimic and influence behaviour as well as traditional techniques such as sounds and pheromones is a great addition to researchers‘ toolkit, says Terry Page at Vanderbilt University in Nashville, Tennessee.

Ìę

Reference:

Here’s a riddle: 100 prisoners stand in line, one in front of the other. Each wears either a red hat or a blue hat. Every prisoner can see the hats of the people in front – but not their own hat, or the hats worn by anyone behind. Starting at the back of the line, a prison guard asks each prisoner the colour of their hat. If they answer correctly, they will be pardoned. Before lining up, the prisoners confer on a strategy to help them. ?

Stumped? Google’s artificial intelligence wasn’t. A deep neural network based on one built by Google’s DeepMind team – software that has previously taught itself to play several dozen classic arcade games – can now solve riddles like this one. The conundrum is reportedly among those posed to prospective .

The ability could one day help groups of robots work together to solve puzzles by sharing pieces of information. “It’s basically a first step toward having AIs that can communicate and collaborate,” says Jakob Foerster at the University of Oxford, who collaborated with Google’s DeepMind team on the research. “In the long run it will give them a lot more scalability and allow them to solve tasks that previously weren’t possible.”

To solve the hat riddle, each person was modelled as a separate artificially intelligent agent. These considered the colour of the hats they could see, decided what to tell the others, and then used the information collectively to work out the answer.

As when learning to play video games, the AIs largely worked out how to tackle the problem themselves. “They’ve come up with protocols that are different from how humans solve these problems,” says Foerster. “We don’t yet fully understand what the solutions are, but we know that they work.”

Paper source:

Like this? Explore more: How computers will be the mother of invention