In a lagoon in Venice, robotic lily pads float on the surface, with clusters of electronic mussels resting on the bed below. Meanwhile a robotic fish swims between them exchanging messages.
This isn鈥檛 a scene from a watery sci-fi, but instead the setup of an experiment into environmental monitoring and robot culture.
In July, a self-organising team of robots will be released into the murky waters of a lagoon near Venice, Italy. Their main job will be to keep an eye on anoxia 鈥� a phenomenon where water loses its dissolved oxygen, which kills fish and other marine creatures.
Advertisement
Anoxia can be onset by natural causes such as tidal disturbances, but can also occur as a result of pollution or sewage dumping. It is unpredictable and difficult to track.
Sitting on the surface will be three aPads, devices that resemble lily pads carrying solar cells, and twelve aMussels, sensors which fall to the lagoon bed, gathering data on things like oxygen levels, salinity, and microscopic lifeforms. In total the team will monitor an area two hundred metres across.
When an aMussel detects something interesting, or its battery runs low, it will bob to the surface and summon an aPad. The two dock and the aMussel recharges and transfers its data to the aPad, which is then sent back to base. Radio does not work underwater, so this little exchange allows the aMussels to communicate what they鈥檝e found.
The aPad can then take the aMussel to a new location if needed. The team will also trial a robotic fish that will swim between clusters of aMussels transferring data.
The plan is to share learned information. For example, an aPad could learn how to use currents to keep to a straight line and this knowledge is then passed on to aMussels to form a sort of collective memory. Over the long term, different clusters should build up 鈥榗ultures鈥� based on local conditions.
Thomas Schmickl, who leads the project at the Artificial Life laboratory of University of Graz in Austria, wants robot swarms to do useful work rather than being confined to laboratories. His project鈥檚 ultimate aim is to develop robot societies in which information is shared and persists after individual robots are replaced. The robots are designed for mass production, with the aim of much bigger swarms in future.
鈥淚f the subCULTron approach can be scaled to monitor environments for many months at a time and adapted to more extreme environments, such as open-ocean, there is enormous potential for this system,鈥� says Catherine Harris who works with autonomous systems at the National Oceanography Centre in Southampton.
