Robot guides reveal A+ tutor-ant technique

Abstract

A good teacher can make all the difference: inspiring, educating and changing lives. For tiny rock ants (Temnothorax albipennis) that live in colonies of about 100 workers, good teachers can be essential for the survival of the community. When a homeless colony relocates, scouts scurry about in search of the next residence. However, instead of laying down a pheromone trail for other members of the colony to follow to their new home, the scouts indulge in a spot of one-on-one tuition, known as a tandem run, guiding individual novices and pausing repeatedly to allow their acolytes to gain their bearings and learn the route. ‘[It] probably reduces the risk that individuals become lost’, says Nigel Franks from the University of Bristol, UK. But he was curious to understand what makes a good ant tutor; ‘what elements are necessary and sufficient for successful teaching’, he says. However, as they couldn't ask the tiny ant instructors to reveal their secrets, Franks and colleagues from the University of Bristol decided to find out how well novice ants learned when guided by a robot instructor. Alan Worley programmed a gantry above an ant arena to move back and forth with a small sliding robot attached to it, so that Jacob Podesta and Edward Jarvis could direct the robot to move at different speeds along either straight or wavy routes. Next, the researchers placed a rock ant nest in the arena and removed the scent glands from a worker ant, attaching them to the robot to make it smell like an ant teacher as it trained novices. Then the team waited for an intrepid pioneer to emerge from the nest before setting the robot off to guide the novice to a new home 35 cm away, allowing the robot to slow and pause when its apprentice stopped to survey its surroundings. But the true test came when the researchers tracked the student's return trip as it headed back to recruit other members of the nest to follow. How well had the naïve ant explorers learned from their wise robot instructor?Using statistics, Ana Sendova-Franks revealed that the ants that had been directed by the roboguide were more than twice as likely to make it back to their original nest as novices that the researchers had carried to the new location. And when the team compared the success of novice ants that had followed a straight path and those whose robot had meandered, both sets of ants were equally successful at relocating their original nest; the shape of the route made little difference. In fact, the returning ants that had been trained by a wandering teacher tended to take more direct return routes. ‘The sinusoidal paths may constitute longer lessons and thus may give followers more time to learn their surroundings', says Franks, adding ‘It is likely that tandem followers learn both local and more distant landmarks, which could be anything from the walls of the arena to windows in the laboratory’.And when the team released additional ants into the arena after the apprentice and its robot teacher had embarked, the trailing ants followed the wavy trainees’ more circuitous outbound routes. The apprentice ants were leaving behind pheromone trails for the trailer ants to follow. Franks explains that teacher ants usually deposit a pheromone trail when guiding novices, but it appears that apprentices can also leave such trails; ‘They may serve as safety lines to help ensure that individuals do not become lost’. And he is keen to find out whether trainee ants can, in turn, influence their robot tutor.