Pavement inspection is one of the most important tasks for the maintenance and rehabilitation (M&R) of transportation road surfaces. There are four shortcomings, as follows, in current pavement inspection systems for project-level inspection: time consuming and labor intensive, inefficient and inaccurate, and require post-processing. In order to deal with these drawbacks, this research developed and designed a real-time autonomous pavement inspection robot, called the PI-bot. We also developed the innovative Chromatic Dual-Light Inspection (CDLI) method, integrated with the PI-bot, which can provide distress-enhanced and spillage-removed effects to find distresses. For experiments, we acquired 504 images and tested the performance of the PI-bot and CDLI method. The results show that the CDLI method performs well for normal pavements, alligator cracks, and manholes, and the performance of CDLI in identifying spillage pavements and longitudinal and transverse cracks is also very successful. This research demonstrates that this real-time pavement inspection system is workable and has high potential for pavement inspection tasks. This system is expected to reduce manpower and cost for pavement inspection applications.