The field of road pavement engineering has seen an explosion of artificial intelligence–based applications since the late 1980s. Such applications are found at key stages of the decision process involved in pavement management. In the analysis phase, they perform pavement diagnosis and deterioration modeling tasks. In the design phase, they enable rehabilitation needs to be assessed and contribute to the identification and selection of maintenance actions. Finally, in the choice phase, they are applied for priority programming of rehabilitation and maintenance. They either represent alternative approaches to existing systems or collaborate to make the overall system more efficient. The purpose of this article is to summarize the findings of up-to-date research articles concerning the application of artificial intelligence to pavement management and to illustrate the potential such tools can offer to pavement engineers. Artificial intelligence techniques include expert systems, artificial neural networks, fuzzy logic, genetic algorithms, and hybrid systems.

Artificial intelligence--based decision support technologies in pavement management

Artificial intelligence techniques represent an efficacious approach both to implement predictive models and to resolve optimization problems. They are, therefore, particularly appropriate for pavement mainetnance management. In this paper, a procedure has been defined to make use of the available economic resources in the best way possible for resurfacing interventions on flexible pavements by using artificial neural networks and genetic algorithms. The neural networks, in particular, were utilized to define both a Sideway Force Coefficient prediction model and an accident prediction model. The optimization problem was resolved by means of an opportunely defined genetic algorithm using the results of the neural networks designed. The procedure was applied to the A18 motorway of the Eastern Sicily road network. The obtained results have highlighted that the procedure represents an efficient approach to obtain one of optimal solutions in a very big space of possible solutions in sufficiently short periods ...

A model based on artificial neural networks and genetic algorithms for pavement maintenance management

Practical optimization of infrastructure preservation works programming has always posed a computational challenge due to the complexity and scale of the problem. Critical to the process is formulations in which the identity of individual projects is preserved. This requirement leads to exponential growth of solution space, often resulting in an unmanageable process using traditional analytical optimization techniques. In this paper, we propose an evolutionary-based multiyear optimization procedure for solving network level infrastructure works programming problems using a relatively new concept known as the shuffled complex evolution algorithm. A case study problem is analyzed to illustrate the robustness of the technique. The findings show convergence characteristics of the solution and demonstrate that the algorithm is very efficient and consistent in simultaneous consideration of the trade-off among various infrastructure preservation strategies. It is concluded that the robust search capability of the shuffled complex evolution technique is well suited for solving the combinatorial problems in network level infrastructure preservation works programming.

Shuffled complex evolution algorithms in infrastructure works programming

Maintenance management of a large road network with limited resources is a challenging task in developing countries like India. Road agencies expect that the condition of their road infrastructure should always above the desired level of performance even under restricted resources. This requires an integrated decision support system considering all aspects of maintenance. In this paper, a framework for optimal maintenance decisions by multi-objective approach is formulated for a road network by considering functional condition of the pavement quantified in terms of roughness and structural condition quantified in terms of rebound deflection. A multiobjective optimization model is developed using Integer Linear Programming (ILP). A novel approach to track the age of pavement is applied in the formulation of the mathematical model. The e-constraint method is adopted to generate non-dominated solutions. Budget bound optimizing model is formulated and implemented for a typical road network of four roads and optimal maintenance scheduling is arrived.

Decision Support System for the Maintenance Management of Road Network Considering Multi-Criteria

Transportation officials are faced every day with competing investment demands, and they must distribute limited resources in the presen ce of uncertainty or in the absence of complete and accurate information. Therefore, they must have access to valid, efficient, and robust asset management tools, which can support the optimization of maintenance interventions and support the policies at all decisional levels. Within this context, emerging technologies, such as artificial intelligence techniques, provide efficient alternatives to the traditional mathematical approaches. In this paper a hybrid approach for optimal management of road pavem ent maintenance for a pluri-annual planning period is proposed. The proposal of the procedure has been the optimal management of the rehabilitation interventions of pavement friction. In particular, a Sideway Force Coefficient prediction model has been cre ated through an appropriate neural network to characterize the pavement conditions. Defined this model it was possible to solve the optimization problem. A genetic algorithm was used to tackle the combinatorial nature of the network -level maintenance programming. The procedure was applied to a motorway infrastructure belonging to the motorway network of Eastern Sicily (Italy).

A Hybrid Approach for Pavement Maintenance Management of an Italian Motorway

The realization of full advantages of pavement management system (PMS) implementation at the network level requires that an optimization analysis be carried out to achieve an optimal programming of pavement activities over a multi-year planning horizon. This paper describes the development of a computer model based on genetic algorithms to solve the network level pavement management programming problem. Numerical examples are presented to illustrate the applications of the program to obtain solutions for three pavement management problems involving different pavement network objective functions and system parameters. Through these examples, the potential uses of the model are highlighted.

T F Fwa

Papers

Network level programming for pavement management using genetic algorithms