NHAI launches AI system to improve highway project quality and efficiency

NHAI Deploys AI System to Address Faulty DPRs and Road Defects

In a significant move to enhance the quality of highway projects and streamline operations, the National Highways Authority of India (NHAI) has deployed an in-house artificial intelligence (AI) system. This system is designed to identify deficiencies in Detailed Project Reports (DPRs), analyze road defects, flag faulty signage, and draw insights from past arbitration cases and project delays. Unlike other AI tools, this system operates exclusively on NHAI's extensive repository of internal documents, including circulars, acts, IRC codes, and monthly reports, without relying on external internet sources.

Currently, the AI system is accessible only to NHAI officials, serving as a tool to reduce the workload of navigating through hundreds of pages of guidelines, standard operating procedures (SOPs), and codes. However, NHAI Chairman Santosh Kumar Yadav announced plans to extend access to consultants and contractors after ensuring safeguards against potential misuse of sensitive information.

Explaining the rationale behind the initiative, Yadav said, "This system was needed because when we conducted a workflow study of Project Directors and regional officers last year, we found they were burdened with tasks that could be automated. With this system, any official can upload a photo of a highway defect, and the AI will analyze it against hundreds of guidelines and provide a result."

A key focus of the AI system is addressing the issue of substandard DPRs, which form the foundation of highway projects. Union Minister of Road Transport and Highways Nitin Gadkari has frequently highlighted the role of poor DPRs in increasing road accidents and fatalities. To tackle this, the AI system allows users to upload a DPR for review. It then cross-references the document against IRC codes, Ministry of Road Transport and Highways (MoRTH) guidelines, and standard policy circulars to identify errors. This proactive approach aims to rectify issues before project execution.

The AI system includes a specialized tool called the Technical Schedule Analyzer, which evaluates the Schedule B and Schedule C annexures of DPRs. These annexures outline the precise physical scope of highway projects. The tool flags issues based on severity and assigns a completeness score, enabling early detection and resolution of potential problems.

Another feature of the system is Margsarthi, an AI-powered chat agent that assists officials with real-time queries. For instance, a field engineer can upload a site photograph and ask Margsarthi for guidance based on existing guidelines. The system, which is integrated with NHAI's datalake, maps the entire lifecycle of road construction projects. Since its launch on April 18, it has processed over 50,000 queries from approximately 1,100 users, with over 40% of these related to circulars and documents.

A senior engineer involved in the project highlighted the system's utility, noting that it addresses the challenge of managing the vast volume of rules and regulations. Previously, officials often relied on personal folders of reference materials, which made it difficult to stay updated. The AI system now enables them to retrieve relevant information with a single query.

In addition to these tools, the AI system includes Kick-Off and MPR Insights, which monitor issues raised during kick-off meetings and monthly project reports. These tools flag critical and overdue issues, provide recommended actions, and link directly to the original source documents, ensuring efficient issue resolution.

By leveraging advanced AI capabilities, NHAI aims to improve project efficiency, enhance safety, and ensure the timely delivery of high-quality highway infrastructure.