This paper describes autonomous operation planning system that can plan the cutting operations rapidly with genetic algorithm. An operation that has two processes, a rough process and a finish one, is examined to determine cutting conditions with evaluating machining cost and tardiness time. Adaptive prediction in the planning allows us to estimate machining cost and processing time with predicting cutting force, tool wear, and surface roughness. Penalty for tardiness, then, can be estimated using the processing time. The optimum combination of cutting conditions in two processes can be searched in the following way : (1) For giving initial conditions in the optimization, genetic algorithm finds the optimum cutting conditions independently in each process. (2) Initial conditions are set, around the conditions found to start the search near the optimum conditions for all process. (3) Genetic algorithm finds the optimum conditions to minimize machining cost including the penalty. It is, then, shown that the planning with genetic algorithm can give us the optimum cutting operation rapidly according to cycle time.