Adaptive learning in service operations

We propose a decision analytics approach that leverages adaptive learning in the refinement of service operations. We aim to integrate service design and service pricing with downstream operational decision-making related to service provision. This approach involves: collecting consumer data and establishing consumer behavior models; integrating consumer behavior models with models for service operation decision-making; and iteratively evaluating service designs based on service delivery performance that evolves over time due to learning. We discuss how this approach enables service providers to set time-differentiated prices and evaluate the impact on transportation network performance. We use agent-based simulation to illustrate the application of our approach to the operations of a public rail transportation firm in a European urban setting. Our findings suggest that knowing the impacts of consumer responses in service operations is essential for devising cost-effective and value-bearing service designs. Our approach can support service providers who wish to adjust their pricing, consumer demand and capacity management models, and to develop more effective market forecasts of performance through adaptive learning, in the presence of “big data” from consumers and operations.