Font Size: 

The current development of the road network is rapidly growing, including the development of the district road network as an infrastructure to improve the regional economic. This is also the case for the development of the arterial road network in Metro City Lampung-Indonesia as the selected location of the current study. Frequent and recurrent road damage is a problem for the local government, therefore a deep evaluation of road deterioration is required to ensure proper selection of corrections. The evaluation of pavement performance using pavement condition indicators is a basic component of any pavement management system (PMS). In this sense, pavement condition index (PCI) have been commonly used to assign a maintenance strategy for the existing pavements. However, a conventional PCI approach either need more steps or expensive calculation software. Therefore, this study aims to develop an alternative simple method by using optimization techniques of back propagation artificial neural networks (ANN). For this purpose, more than 22 sections of roads were investigated and the results showed that the most dominant distresses are alligator cracks and rutting that require more handling. The calculation performed by ANN has shown that the priority of care should be provided in section 15 with the smallest value of 51.23. This result is comparable with that of conventional PCI method which is 51.42. It can be concluded that ANN soft computing can predict the road damage in a simpler procedure.