Adaptive Systems—A Case for Calculating Estimated Time of Arrival

In this paper a system that can estimate the time of arrival for scheduled transport is presented. In the case presented here, the system is applied to a regular boat service. It has been in use for a six-month period, during which it has been able to give very accurate ETA values for both normal and exceptional situations. An advantage with learning systems is that they are general and can be used for new applications without extensive setup.