Reverse Engineering of Maintenance Budget Allocation Using Decision Tree Analysis for Data-Driven Highway Network Management

Abstract

One important aspect of network-level highway management is the rational distribution of the maintenance budget to the necessary assets. However, the decision making underlying budget allocation is often unclear, making it difficult to determine whether the budget is being allocated effectively. Based on the PDCA (plan–do–check–action) approach to maintenance management, this research proposes the application of decision tree algorithm to reverse engineer the factors affecting maintenance budget allocation. Annual inspection and budget data for 3000 km of highway network were analyzed using the CART algorithm with two conceptualizations of budget allocation. Both frameworks revealed that the budget allocation was related to factors other than pavement conditions, and it was concluded that maintenance planning was primarily based on subjective considerations, rather than inspection data. This study demonstrates the combination of PDCA cycle and decision tree analysis as a valuable technique for evaluating and improving decision making in maintenance budget allocation and highway network management.