Simulated annealing was invented in 1983 by Kirkpatrick, Gelatt, and Vecchi.
It was first used for placing VLSI* components on a circuit board
Simulated annealing has also been used to solve the traveling salesman problem, although this approach has proved to be less efficient than using heuristic methods that know more about the problem.
It has been used much more successfully in scheduling problems and other large combinatorial problems where values need to be assigned to a large number of variables to maximize (or minimize) some function of those variables.