Thinking back on how ASSISIbf began, I remember a white paper napkin of the first sketch of a robot talking to a honeybee I was sketching together with some fellow scientists. I remember how excited I was this might actually happen one day, to have a robot that teaches itself how to communicate with animals.
Soon, another idea came up: to introduce robots not only into a honeybee society, but to think of another robot type, able to interact with a swarm of fish. The idea was to find common traits in the social interaction of both species.
How could we insert the robots so deeply into the swarms that they could modulate their collective behaviors from within? If we managed and understood how to do it, by applying it to two very different animal swarm species, we might find common aspects of animal collective decision-making in general, applicable across all different collective decision-making systems.
That was the idea of ASSISIbf. A simple, but challenging one. We built a consortium of experts and drafted a proposal, and nursed it to the end to happily see it funded by FET (Future and Emerging Technology) in the FOCAS call.
In 2013, we kicked-off the project and started our research activities.
Robots ensure bees get the buzz - ASSISIbf on the Euronews Futuris video series
Now, almost six years later, I realize how ambitious our plan was. As always, ideas sound simple at first, but then turn to become complex quickly: it took us years to build robots sophisticated enough to be accepted by real animals.
We used evolutionary computation and machine learning to allow the robots to learn by themselves how to "speak" to the animals. Driven by mathematical models based on feedback loops of animal swarm behavior, we did it, we finally managed to have our robots acting as "benevolent agents provocateurs" in both animal societies, and to meddle with their collective decision making in order to study and understand them.
As a team we finally were able to realize the dream of creating adaptive biohybrid societies - societies that consist of living animals, mechatronic robots and computer algorithms that govern both of them as a collective. But this is not the end of the story.
After having created those "social cyborgs", as we call them, we set on for another challenge, far beyond just studying swarms from within..this is another story, one that still has to stay a secret for some time. Thus, the full story of ASSISIbf will be told in a later post.