An iterative repair search method is described called constraint based simulated annealing. Simulated annealing is a hill climbing search technique capable of escaping local minima. The utility of the constraint based framework is shown by comparing search performance with and without the constraint framework on a suite of randomly generated problems. Results are also shown of applying the technique to the NASA Space Shuttle ground processing problem. These experiments show that the search methods scales to complex, real world problems and reflects interesting anytime behavior. Zweben, Monte and Davis, Eugene and Deale, Michael Ames Research Center ...