An amazing application of a colony animals solution to a search for food could revolutionise the way robots communicate.
Israeli researcher Yossi Yovel leads the team at Bat Lab for Neuro Ecology at Tel Aviv University. His research into communication has focussed specifically on the Mexican fish eating bat, native to California. The large range these bats have to cover at night in order to find food has led to them evolving an innovative approach to catching the small fish at the surface of the water.
By working in small groups, they are able to effectively spread their search area, yet still pick up sounds from the other bats. By hunting solely, but remaining in contact with the others, a bat can eavesdrop and track the success of the colony. Yovel addressed the YPO Edge Conference in Singapore last week, explaining how the researchers affixed small sensors to the bats which in turn could formulate an algorithm from flight studies. This algorithm could be used to help robots work together more efficiently by employing similar sound based communication.
Bats employ echolocation to communicate and effectively visualise their night time environment, often to find food. By making sounds and waiting for the returning echo, they can build a sonic picture of their environment, gauging distance by how long it takes a sound to return.
The potential application for robot teams is huge: topography, environmental clean-up – the list is endless. A small group of robots is being built by Yovel and his team; each bot would have a speaker and twin microphones to emit and receive sound respectively.
Yovel explained these would initially be deployed underwater and use the algorithm generated from the fish-eating bats echolocation hunting applied as sonar. The robot collective can locate objects individually at the same time as learning of the others’ findings.
In its promising stages, this research could come to fruition in the next 5 years where we could potentially see a group of robots trained to use purely sound for communication. Yovel wants to prove the application of animal communication systems to modern robot design, there’s no real practical implementation at this stage. The research is a testbed for the algorithms workability. Should someone show an interest a more specialised robot can be developed for any project task.