Heuristically use domain knowledge to transform an initial safety problem into a global optimization problem.
When function evaluation is costly, information-based optimization makes good use of such information to reduce computational costs.
In summary, one can transform an initial safety problem into a global optimization problem through heuristic use of knowledge of one's problem domain. Since such optimization involves a simulation for each function evaluation, we desire global optimization methods which make good use of costly function information gained in the course of optimization. Information-based approaches which approximate mathematically complex decision procedures are most promising for such applications.