Blog: Smart Swarms of Bacteria Inspire Robotics Researchers

Smart Swarms of Bacteria Inspire Robotics Researchers
American Friends of Tel Aviv University (11/17/11)

Tel Aviv University (TAU) researchers have developed a computational model that describes how bacteria move in a swarm, a discovery they say could be applied to computers, artificial intelligence, and robotics. The model shows how bacteria collectively gather information about their environment and find an optimal plan for growth. The research could enable scientists to design smart robots that can form intelligent swarms, help in the development of medical micro-robots, or de-code social network systems to find information on consumer preferences. "When an individual bacterium finds a more beneficial path, it pays less attention to the signals from the other cells, [and] since each of the cells adopts the same strategy, the group as a whole is able to find an optimal trajectory in an extremely complex terrain," says TAU Ph.D. student Adi Shklarsh. The model shows how a swarm can perform optimally with only simple computational abilities and short term memory, Shklarsh says. He notes that understanding the secrets of bacteria swarms can provide crucial hints toward the design of robots that are programmed to perform adjustable interactions without needing as much data or memory.

View Full Article

Blog: Army of Smartphone Chips Could Emulate the Human Brain

Army of Smartphone Chips Could Emulate the Human Brain
New Scientist (05/04/10) Marks, Paul

University of Manchester computer scientist Steve Furber wants to build a silicon-based brain that contains one billion neurons. "We're using bog-standard, off-the-shelf processors of fairly modest performance," Furber says. The silicon brain, called Spiking Neural Network Architecture (Spinnaker), is based on a processor Furber helped design in 1987. Spinnaker's chips contain 20 ARM processor cores, each modeling 1,000 neurons. With 20,000 neurons per chip, Furber needs 50,000 chips to reach his goal of one billion neurons. A memory chip next to each processor stores the changing synaptic weights as numbers that represent the importance of a given connection. As the system becomes more developed, the only computer able to compute the connections will be the machine itself, Furber says. Spinnaker relies on a controller to direct spike traffic, similar to a router for the Internet. The researchers have built a small version of the silicon brain with 50 neurons and have created a virtual environment in which Spinnaker controls a Pac-Man-like program that learns to find a virtual doughnut.

View Full Article

Blog: They're Robots, but Not as We Know Them; the co-evolution of the brain and body in robots

They're Robots, but Not as We Know Them
Computerworld New Zealand (12/02/08) Hedquist, Ulrika

Neural networking was the focus of the 15th International Conference on Neuro-Information Processing in Auckland. Researchers discussed how a better understanding of the brain could lead to more intelligent computer systems. According to Nik Kasabov, director of the Knowledge Engineering and Discovery Research Institute at AUT University, neuro-information processing has real-life applications in medicine, cybersecurity, and intelligent robots. Researchers from the Okinawa Institute of Science and Technology in Japan showed off neuro-genetic robots. "Robots are now not only based on fixed rules about how to behave, they now have genes, similar to human genes, which affect their behavior, development and learning," said Kasabov. And researchers from the German Honda Research Institutes discussed the co-evolution of the brain and body in robots, and robots that can change their shape were also on display. "They can evolve, in a similar way as [humans] evolve," said Kasabov.

View Full Article

Research: Language of a Fly Proves Surprising; research may improve neural networks

Language of a Fly Proves Surprising
Los Alamos National Laboratory News (03/10/08) Rickman, James E.

Researchers have developed a way to view the world through the eyes of a fly and partially decode the insect's reactions to changes in the world around it. The research has changed scientists' understanding of neural networks and could provide the basis for intelligent computers that mimic biological processes. The researchers used tiny electrodes to tap into motion-sensitive neurons in the visual system of a blowfly. The fly was harnessed into a turntable-like mechanism that mimicked the kind of flight it might undergo when evading a predator or chasing another fly. The neurons' firing patterns were mapped with a binary code of ones and zeroes. The researchers found that the impulses were like a primitive, but very regular "language," with the neurons firing at precise times depending on what the fly's visual sensors were trying to tell it about its visual stimulus. Previous research showed irregular spikes in the neurons' firing, but this is now believed to be a way to conserve energy when there is little change in the fly's surroundings. The simulated flight creates significant change requiring regular neuron firing to process the information. "This may be one of the main reasons why artificial neural networks do not perform anywhere comparable to a mammalian visual brain," says Los Alamos physicist Ilya Nemenman, a member of the research team. The research could improve the analyses of satellite images and facial-pattern recognition.
Click Here to View Full Article

Research: A New Theory Changes the Thinking Behind Creating Robots and Smart Machines

A New Theory Changes the Thinking Behind Creating Robots and Smart Machines
Knowledge@W.P. Carey (02/13/08)
The school of Connectionism postulates that the human brain learns when neurons link experiences and understandings, and that the development of artificial intelligence hinges on emulating this capability with computers. But W.P. Carey School of Business professor Asim Roy has challenged these long-cherished notions in an academic paper where he argues that while connections between neurons are necessary, the system still requires organization by a controller. Roy presents a theory that elements of the brain are controlled by other elements, and has partly validated it by demonstrating that Connectionist brain-like learning systems are guided by higher-level controllers, in defiance of the Connectionist view that they employ only local controllers at the neuron level. "What I did was structurally analyze Connectionist algorithms to prove that they actually use control theoretic notions even though they deny it," says Roy, adding that he used neuroscientific evidence to support his argument. The design of various types of robots will eventually be affected by the rethinking of human learning and brain function that Roy's paper has engendered. Roy cautions, however, that his theory may not effectively change computer operations for decades.
Click Here to View Full Article