Discrete-order admission ATP model with joint effect of margin and order size in a MTO environment

Swarming demands and seasonality periodically induce a stringent capacity problem in the made-to-order (MTO) B2B environment. Desirable order admission is a tactic with minimal application cost for handling this problem. A real-time order admission problem with limited major-customer population and batch-size demand with heterogeneous distributions is modeled as a dynamic and stochastic knapsack problem. The optimality of the Markovian deterministic reward-threshold policy is verified. Optimal policy, basic policy, and the capacity rationing method are compared in the designed experiment. The numerical results suggest great potential for increasing profit using the optimal order admission policy for MTO businesses.