Professor Hoon Sohn’s research group has developed an automated inspection system that combines infrared and visual sensing to objectively evaluate steel bridge paint. The system measures coating thickness, detects damage, and assigns condition ratings, improving inspection safety, consistency, and maintenance decision-making.

Paint is one of the important protective components on steel bridges, yet its condition is difficult to assess reliably. By shielding steel from moisture, salt, and air, paint helps prevent corrosion and extends the service life of bridges. When paint degrades, defects such as corrosion can develop beneath the surface before becoming visible, threatening both structural safety and maintenance efficiency. Thus, regular inspection of bridge paint condition is essential.

Despite its importance, paint inspection remains largely manual, typically relying on visual observation and contact-based measurements at a limited number of points. These methods are time-consuming, potentially dangerous often requiring work at height or beneath bridge decks, and subjective. Results can vary depending on the inspector experience and site accessibility. As infrastructure ages and maintenance demands increase, these limitations clearly highlight the need for a safer and more objective inspection approach.

To address these challenges, Professor Hoon Sohn’s research team in the Department of Civil and Environmental Engineering at KAIST developed an automated system to evaluate steel bridge paint condition in a consistent and data-based manner. Rather than relying solely on human judgment, the system combines active infrared-based sensing and visual observation to inspect large surface areas while minimizing direct human involvement.

The system works by heating the painted surface as it moves along the bridge and observing how the surface responds. An infrared camera captures temperature changes that reveal variations in paint thickness and presence of defects, while a RGB camera captures surface images showing visible characteristics on surface. By analyzing these two types of information together, the system can evaluate visible and invisible paint conditions across an entire inspection area.

In the evaluation process, paint thickness is quantified over the entire area, damaged areas are identified and measured, and these findings are combined to produce an overall paint condition rating (Fig. 1). This process is designed to follow the existing inspection guidelines, allowing engineers and maintenance planners to interpret results easily and consistently. 

Figure 1 Paint condition assessment process

The system provides not only the numerical result, but also the visualized results. Paint thickness variations are visualized across target surfaces and the identified and classified defects are also shown with the location of each defect (Fig. 2). A. minute thickness variation over than 20 mm is successfully revealed and early-stage or hidden defects can be detected before serious damage occurs.

Figure 2 Paint thickness assessment results for defect specimens: (a) corrosion specimen, (b) delamination specimen, (c) checking specimen, (d) chalking specimen

Another advantage is the system’s ability to operate in areas that are difficult or hazardous for humans to access. These include bridge sides and undersides, where inspections often involve safety risks. The system can successfully operate on the steel bridge, an area that would normally require risky manual inspection (Fig. 3). Field tests confirmed that the system could safely navigate and inspect such locations while delivering stable and repeatable results.

Figure 3 Lateral operation test: (a) automated paint condition assessment (APCA) system at the starting location and (b) inspection path of sideways.

By transforming paint inspection into an objective and repeatable process, this technology enables safer and more efficient bridge maintenance. Automated evaluation reduces inspection risks and supports consistent, data-driven maintenance decisions.