This is where the fruit fly comes in. When a fruit fly is developing, its back needs some cells to develop into its exoskeleton and some to develop into sensory bristles. Too many bristles, and the skeleton isn't strong enough; too much exoskeleton and the fly is ill-equipped to sense its environment. But there's not a central system that dictates which cells will develop into exoskeleton and which into bristles. Each cell holds the potential for both, and when a cell starts developing one way, it sends a chemical message to its neighbours. A bristle developing in one cell will tend to suppress bristle development in its neighbors, so equilibrium is established, Shackleton said. This decentralised model, in which each cell or base station settles with its partners, works in Shackleton's tests. "It will come up with a useful plan which minimises interference," and can better adjust to changing usage patterns, he said. Which antennas use which frequencies would no longer be BT's problem. The antennas could simply work it out among themselves. Shackleton said BT is also applying the work to military communications systems, which have even worse problems because the equivalent of base stations are mobile and the chaos of military actions makes it hard to keep track of a network's components. Shackleton also described a threat-detection system in which computers on a network occasionally check their neighbours' status. When a computer has an unresponsive neighbor, its own status changes and it goes on alert, which in turn sends other computers into a cautious state. Such a system can help reconfigure security systems to wall off intrusions as they spread, Shackleton said. Loosening the reins
Letting elements of a computing system autonomously govern themselves in this manner has advantages, but it's likely to ruffle feathers in traditional information technology quarters, where people are used to being in control, Shackleton said. "We must be willing to give up a certain amount of control, at least of detailed control," in order to let these self-regulating systems succeed, Shackleton said. And as a related matter, it won't always be clear exactly how independent entities arrange themselves to arrive at a workable configuration. This loss of comprehensibility to human administrators is "possibly the most contentious (issue) of all," Shackleton said. Another consequence could be a loss of performance, in which self-governing systems work well enough, but not as well as systems run by human engineers, Shackleton said. But Steve White, senior manager of IBM's autonomous computing division at IBM's Thomas J. Watson Research Center, said lower performance shouldn't be an option. "I think we should be shooting for autonomous self-administering systems that can do better than humans do," Watson said in an interview.





