Teams of robots that can remotely tap into each other鈥檚 sensors and computers in order to perform tricky tasks have been developed by researchers in Sweden. The robots can, for example, negotiate their way past awkward obstacles by relaying different viewpoints to one another.
Robert Lundh, who developed the bots at 脰rebro University, says cooperative behaviour is normally rigidly pre-programmed into robots. 鈥淲e wanted to have the robots plan for themselves how to draw on their capabilities and those of others,鈥� he told New 杏吧原创.
Lundh鈥檚 robots decide whether another nearby robot may be able to help with a specific task. In one experiment two round robots, each 45 centimetres in diameter and 25 cm tall, teamed up to negotiate their way through a doorway. They were forced to cooperate because each robot鈥檚 vision system had been limited so that it could not see enough of the doorway to be certain of getting through without hitting the sides.
Advertisement
Stereo vision
The first robot determined that it needed to wirelessly access the camera of its partner and then navigated through the doorway using both its own camera and that of its colleague. Once through the door, it turned around and returned the favour, helping the second one through by providing a wider field of view.
鈥淥ur system allows robots to start with a task, extract which capabilities are needed and find out where to access them,鈥� Lundh explains. 鈥淚f you don鈥檛 have the capabilities on your own you have to search for them.鈥�
Another experiment involved a pair of robots carrying one length of wood balanced on each of their heads. One of the robots tapped into information about the other鈥檚 speed and direction to keep the beam balanced correctly.
Flexible teams
The strength of the system, says Lundh, is that the robots can adapt their teamwork to the task in hand, and to the availability of potential team mates. 鈥淗umans don鈥檛 exist in fixed teams,鈥� he says. 鈥淭hey cooperate as needed.鈥�
Ken Young, a robotics expert at Warwick University in the UK, believes it makes sense to model robot cooperation on that of humans. He adds that teams of small robots should be more effective than large individual ones. 鈥淗umans regularly work together and combine skills to perform tasks not possible alone,鈥� he told New 杏吧原创. 鈥淭hat鈥檚 an obvious route forward for robots as well.鈥�
Lundh hopes to test the new system within an 鈥渋ntelligent home鈥� filled with cameras and radio frequency tags that mark specific locations. The bots could then tap into these systems as well.
