In the wild, animals affect each others behavior all the time—prey species keep away from predators’ turf, insects jump away from browsing herbivores—they may not be communicating per se but they shape each others’ environment. To further explore this relationship, researchers used robots to mediate between two species unlikely to ever encounter each other in the wild: zebrafish and honey bees. The researchers chose these species because they are both social and gregarious animals who regularly make decisions as a group. To do these experiments the animals were kept apart– in fact, they are in different countries. In Switzerland, a small school of zebrafish swim around a circular path, with a camera overhead tracking their position. Swimming along with them is a robot lure, the same size and shape as a zebrafish. Meanwhile in Austria, a swarm of young bees buzz around an oblong enclosure, with two robots in the middle. The bee-bots generate heat, which causes the bees to congregate around them. A proximity detector tracks the bees’ location. In each setup, the fish and bees have a binary collective choice to make – the bees can choose to congregate around one of two robots and the fish can swim either clockwise or counterclockwise around the circle. The researchers wanted to know if, using robots as mediators, the two species could come to a collective decision – could the species agree which direction to swim in and which bee-robot to gather around? Four different experiments were performed with varying degrees of communication between the animal groups. The first experiment established a control group, with no communication between the fish and the bees. In the second experiment, the bees’ behavior was observed and then transmitted to the fish robot, affecting which way it lured the school of zebrafish. After six minutes, the bees began to congregate around the left bee robot, causing the fish robot and, as a result, the fish to swim in a counter clockwise direction. In the third experiment, the fish’ behavior was observed and then transmitted to the bee robots, affecting which robot produced heat. The fish constantly changed the direction in which they were swimming, which caused the bees to continuously switch sides as well. In the final experiment, the behavior of both the bees and the fish was observed, and transmitted both ways At first, the fish and bees couldn’t come to a decision– one side did one thing, the other another, with lots of switching back and forth. But after about 20 minutes, they reached a consensus, with the fish swimming counter-clockwise and the bees congregating around the left bee-robot. This shows that two species, which would have no interaction whatsoever in nature, are actually able to interact and come to a collective decision thanks to the help of the robots. Moving forward, researchers would like to introduce machine learning into these types of scenarios So bee-bots learn behavior from bees and fish-bots from fish, instead of being programmed ahead of time. Researchers believe that in the future, this technology could be used to mediate species interactions not just in the laboratory, but also in the wild—the research could even be adapted to work with plants, fungi and microorganisms One day, machine-learning robots could be used to perceive a dramatic event in a bee colony, and react to help the bees to find a way to recover. This could be applied to any type of animal group, once the robot knows how they make collective decisions.