Service System Design Under Information Uncertainty: Insights from an M/G/1 Model

We consider the design of a service system in which the service provider performs an operation on incoming transactions for a client. An important task for the service provider is to decide the capacity of the service system based on information available, such as transaction arrival rate, service level required by the client, and fixed/variable costs. In many cases, some input parameters to this capacity planning problem, such as transaction arrival rate, are uncertain at design time, and their values used are estimates. Nonetheless, violating the service-level requirement during actual operation may incur a penalty. We analyze this problem using an M/G/1 queueing model and analytically derive the optimal capacity under the assumption that transaction arrival rate is only known to a range. We use the model solution to obtain some insights on how demand information uncertainty impacts such a service business.