Evolutionary algorithms are being used by the UK's Cranfield University to develop artificial intelligence (AI) for interceptors that would be used in ballistic-missile defence. The algorithms would enable several interceptor missiles, fired at once, to communicate with each other and calculate the best way to strike incoming targets. The AI decisions could see one interceptor engage one target, or multiple missiles hit the same threat. "We would have multiple low-cost interceptors with low-cost sensors instead of one expensive interceptor," says Dr Evan Hughes, a senior lecturer at Cranfield. "However, as they co-operate, the low-cost sensors' performance will be good."
ROB COPPINGER/SWINDON
Source: Flight International