Abstract
The necessity of work-in-process management of rail-cars production by intelligent decision support software, which is based on JIT and Kanban principles, is identified. The scheme of one type cargo-flow movement between two workshops with control by electronic and traditional kanban-card is offered. For simulation of rail-cars production cargo flows Markov chain of M/M/1/1 type was applied. Simulation shows the dependence of the work-in-process on the in-flow and out-flow intensity. To determine the high level of the optimal work-in-process the stochastic inventory management model is applied