SG strengthens asphalt-concrete quality control with AI crack analysis, tracks performance after construction

[Financial News] SG, an asphalt-concrete specialist, said on the 8th that it will strengthen its quality-control system for asphalt pavement using its self-developed artificial intelligence (AI) crack-analysis program.
SG recently conducted AI crack analysis and a pavement performance survey on a test-paved section of “Eco Steel Ascon” at the construction site for Line 5, the fifth semiconductor plant at Samsung Electronics’ Pyeongtaek Campus. Eco Steel Ascon is an eco-friendly, high-durability asphalt concrete made using steelmaking slag.
In the survey, AI was used to analyze images of the pavement surface and assess cracks and damage. The degree of pavement deformation and its ability to withstand loads were also measured.
The company says that while conventional pavement inspections relied mainly on visual checks at the site, AI crack analysis can quantify pavement conditions and continuously track changes in the same section. Accumulated field data can also be used to compare and analyze pavement performance by product and site.
The survey found that the average crack rate in the Eco Steel Ascon section was 0.11%, lower than the 0.93% in the standard asphalt-concrete section. Tests measuring how much the pavement surface had compressed or become depressed also found no deformation in the Eco Steel Ascon section. By contrast, deformation of up to 0.4 millimeters (mm) was detected at some points in the standard asphalt-concrete section.
The elastic modulus, which indicates the pavement’s elastic behavior under load, was measured at about 13% higher for Eco Steel Ascon than for standard asphalt concrete. SG said, “This survey has laid the groundwork for confirming the condition and performance of pavement after construction in numerical terms and tracking changes going forward.”
Starting with the survey of the P5 test-paved section, SG plans to use its AI crack-analysis program for post-delivery monitoring of asphalt-paved sections it supplies, including those paved with Eco Steel Ascon. The company plans to use pavement-condition and performance data accumulated after delivery in quality control, product improvement, and the development of asphalt mix formulations and production technologies.
An SG official said, “This survey confirmed Eco Steel Ascon’s excellent pavement performance using real-world data,” adding, “Going forward, we will use AI crack analysis for post-delivery monitoring to track changes in performance after delivery as well, and actively apply the data we accumulate to quality improvements and research and development.”
[email protected] Choi Du-seon Reporter