A Data-mining Technique for the Planning and Organization of Truck Platoons

In this paper, we introduce the problem of mining frequent sub-routes for the use of platoon driving. The problem is related to the problem of mining frequent sequences. While the problem of mining frequent sequences normally focuses on patterns with a large support, the problem of mining frequent sub-routes has to deal with a small absolute support of two or more. We will present our algorithm to find truck platoon sequential pattern (TPSpan) which is based on pattern-growth methods and solves the problem of mining frequent sub-routes in an efficient way by using so called pruning parameters. We also present experimental results about the amount of possible truck platoons based on synthetic data and about the possibility of finding truck platoons with the use of pruning parameters in context of efficiency, flexibility and reliability.