The problem

In most cities the detection of defects on streets are collected through visual inspections made by operators and through notifications made by residents.

The main limitations of this visual defect detection system are:

  • High time data acquisition. The acquisition speed depends on the skills and speed of the observer.
  • No small defects are detected, only those who are visually perceptible by the observer.
  • Any damage notified by residents is reported when it is hazardous.
  • The observations have a subjective component. The determination of the degree of danger depends on the operator.
  • Excessive time to characterize the defect (measurements and take some pictures of the defect).
  • The defect characterization process requires some affect on traffic to do it.
  • Manual entry of data to pavement management system.
  • Slow process: detection, characterization, alert notification and inclusion in the information system.
  • Process involving the human factor, which can deal into errors.


That is why there is a need for an automatic detection and warning system of asphalt defects that is more accurate and allows a more efficient response.