Evolution by natural selection is Nature's optimization technique. Owing to genetic variation in populations, optimization proceeds in a parallel, rather than a serial, fashion. The advantages of this method have long been recognized in computer science and artificial intelligence, where genetic algorithms simulate the evolution of ‘populations’ of solutions to a problem, with both ‘sexual reproduction’ and ‘mutations’ leading to new solutions whose survival depends on their fitness. Now, genetic algorithms have been successfully applied to two aerospace problems. Geraud Granger and his colleagues at CENA, the Center for Aerial Navigation Studies in Toulouse, France, have applied genetic algorithms to air traffic control. Their program, simulated on actual air traffic data, finds optimal solutions that keep planes from coming too close to each other, whilst at the same time minimizing deviations from the planes’ original trajectories. And at Purdue University, William Crossley and Edwin Williams have applied a genetic algorithm to the complex problem of planning low satellite orbits in order for the group of satellites to cover as much territory as possible. In some cases, the solutions that they have evolved reduce gaps in coverage by 25% – which has surprised satellite engineers. MW