Design a just-in-time organization system using a stochastic gradient algorithm

This paper introduces a new highly fiexible manufacturing organization, seru, which has brought huge benefits to industries. Seru was developed to cope with shorter product life-cycles, uncertainty in product types and demands, and fluctuating production demands (sometimes mass, sometimes batch and sometimes very small volumes). We construct a dynamic seru system model to maximize expected company profit under stochastic production demands and uncertain product types. Because the profit function is not concave, a convergent stochastic gradient algorithm is used to find a stable solution. A numerical example is used to illustrate the proposed stochastic gradient algorithm.