A multi-disciplinary group of scientists has formed the CoralBot collective with the aim of building robots that autonomously repair coral reefs. The systems combine machine vision technology, innovations in computing, marine biology and undersea robotics.
Coral reefs support the livelihoods of almost half a billion people to the tune of 60 billion dollars in ecosystem goods and services like fisheries, coastal protection, tourism and biodiversity. Thus, the need for dealing with the threat of coral reef destruction is quite real, according to the CoralBot collective.
CoralBots are teams of robots that intelligently navigate damaged coral reefs and transplant pieces of healthy corals along the way. Swarm intelligence methods are used to control the robots’ behavior. Swarm intelligence explains how simple behaviors in a group of creatures can lead to complex and functional structures (e.g. how bees build hives, termites build mounds, beavers build dams, and so on), according to CoralBots.
The development and testing of the robots at sea has already been done, and the last task is to embed the robots with computer vision to let them see the healthy bits of coral and to configure the robot arms. Leading the machine vision development for the CoralBot collective is computer vision scientist Dr. Neil Robertson, who leads the vision lab at Heriot-Watt University.
Each of the other team members brings a specific expertise to the group: Dr. Lee-Anne Henry, coral biology; David Lane and Dick Blidberg, autonomous underwater robot technology; and David Corne, swarm intelligence.
In April, the team launched a Kickstarter campaign to raise the funds to purchase and assemble the machine vision component of the program, with hopes of conducting a live demonstration of the CoralBots in a public aquarium. As of May 21, the campaign had raised $31,625 of its $107,000 goal. The campaign is open until May 27.