India's road network is the second largest in the world, spanning over 63 lakh kilometres. At the heart of every durable national highway or state road lies one foundational document: IRC:37 — Guidelines for the Design of Flexible Pavements. First published by the Indian Roads Congress in 1970 and significantly revised in 2012 and again in 2018, this standard governs how engineers calculate layer thicknesses for bituminous roads based on traffic loading, subgrade strength, and material performance.
Yet for decades, compliance with IRC:37 pavement design standards depended on spreadsheets, hand calculations, and post-construction deflection tests — often too late to correct costly errors. Today, AI-based pavement design validation is changing that equation. Real-time AI systems can now cross-check every layer against IRC:37 parameters during both the design and post-construction phases, flagging non-compliance before it becomes a structural failure.
IRC:37 provides the mechanistic-empirical framework for designing flexible pavements in India. It defines:
The 2018 revision incorporated performance-based design requirements, making it essential for engineers working on projects under MoRTH specifications, NHAI highway projects, and Pradhan Mantri Gram Sadak Yojana (PMGSY) roads.
The design begins with assessing subgrade CBR values. IRC:37 pavement thickness design requires a minimum design CBR (usually the 90th percentile value from field tests). A weak subgrade (CBR < 3%) demands thicker sub-base layers, sometimes with soil stabilisation using lime or cement as per IRC:SP:89.
Engineers must calculate the design traffic in MSA using the Vehicle Damage Factor (VDF) and lane distribution factor. IRC:37 uses the fourth-power law to convert mixed traffic into equivalent standard axle loads. Errors at this stage compound through every subsequent layer calculation.
IRC:37 provides pavement design catalogues for standard combinations of CBR, traffic, and material type. For complex projects, the standard permits use of mechanistic-empirical design software such as IITPAVE, which models strain at critical interfaces.
Two critical performance checks govern IRC:37 flexible pavement design:
Manually validating these parameters across dozens of cross-sections on a highway project is time-consuming and error-prone. This is precisely where AI pavement condition monitoring and automated compliance engines deliver transformative value.
The International Roughness Index (IRI) is a key post-construction quality metric under MoRTH specifications for roads. Acceptable IRI thresholds are:
AI-powered road survey vehicles using laser profilometers and computer vision road inspection capture continuous IRI data. The AI engine then instantly benchmarks readings against IRC:37 and MoRTH IRC standards, flagging sections where roughness exceeds permissible limits and generating non-conformance reports automatically.
Rutting is one of the most visible failures in Indian highways, especially in high-traffic corridors like the Delhi–Mumbai Expressway or Chennai–Bengaluru National Highway. AI road condition assessment systems use stereo cameras and structured light scanning to measure rut depth to millimetre accuracy.
When rut depths approach or exceed IRC:37 thresholds (typically 20mm), the AI automatically correlates findings with the original design layer data to determine whether the failure stems from:
This diagnostic capability transforms pavement analysis software from a passive recorder into an active engineering tool for PWD engineers in India
AI systems can process Falling Weight Deflectometer (FWD) data alongside IRC:37 layer design parameters to estimate remaining fatigue life in real time. Machine learning models trained on thousands of Indian highway cross-sections can predict fatigue cracking onset with significantly higher accuracy than traditional extrapolation methods.
During construction, AI-based pavement condition monitoring systems can validate as-laid layer thicknesses using Ground Penetrating Radar (GPR) data, comparing them against the approved IRC:37 design thickness in real time. Deviations trigger immediate alerts to site engineers and QA teams — preventing the silent non-compliance that only surfaces years later as premature failure.
Understanding the application of IRC:37 in Indian highway projects across major infrastructure initiatives is critical:
The next frontier in IRC 37 pavement design validation is integration with Building Information Modelling (BIM) and digital twin platforms. Engineers at NHAI and MoRTH are already piloting systems where every kilometre of highway has a digital twin — a real-time virtual replica that ingests sensor data, compares it against IRC:37 flexible pavement design parameters, and updates maintenance forecasts automatically.
This convergence of AI road inspection technology with established standards like IRC:37 represents the most significant shift in Indian road engineering since the introduction of mechanistic-empirical design.
IRC:37 remains the cornerstone of flexible pavement design in India — but its full potential is only realised when compliance is verified continuously, not just at design stage. AI pavement inspection and automated road survey solutions are now capable of validating IRI, rutting, and fatigue parameters in real time, empowering MoRTH, NHAI, and PWD engineers to build roads that perform as designed, for their full design life.
RoadVision AI's Pavement Condition Intelligence Agent automates compliance checks against IRC:37, MoRTH specifications, and IRC:SP standards — delivering audit-ready reports across your entire road network.
The Autonomous Road Engineer
AI that is transforming road infrastructure, from reactive to predictive.
