PCI assessment through digital images and the use of the Cascade Trainer algorithm to optimize cost and time

The evaluation of flexible pavement condition is essential for decision-making regarding their maintenance and upkeep, given their relevance as an essential component of transportation infrastructure. Consequently, providing adequate maintenance promotes a high level of service, which positively impacts road safety. This study develops a methodology to optimize the evaluation of flexible pavement condition by implementing an algorithm for detecting surface defects using digital images. The proposal aims to optimize resources allocated to preventive maintenance of flexible pavements and promote the development of sustainable road infrastructure. The results obtained show that the implementation of the Cascade Trainer algorithm allows a 49.63% reduction in pavement deterioration level inspection time. Furthermore, a reduction in the total cost of the process equivalent to $965.98 per kilometer was achieved. The evidence obtained highlights that the automation of routine inspections optimizes productivity and maximizes profitability in flexible pavement maintenance.