AI Asked 'What's On Your Mind Today?' Delhi Replied: Detect Potholes And Monitor Road Work

Soon, an Artificial Intelligence-based road monitoring system will detect potholes, debris, and track construction activity in Delhi, as part of efforts to improve the maintenance of city roads.

The Public Works Department (PWD) is working on developing a centralised AI-enabled platform for the aggregation and management of road-related complaints from multiple data sources for efficient tracking and resolution. The government has invited quotations for the development of the system.

The AI-driven system, to be called the Intelligent Road Maintenance and Monitoring System (IRMMS), will detect roadside debris and obstructions using video feeds.

Speaking to News18, a PWD official said that the system shall support the detection of garbage, debris,

and roadside obstructions.

“The detection capability will focus on identifying visible objects that affect road cleanliness, safety, or movement… The system will support timely identification of such issues for field response,” the official added.

They explained that the use of technology will fast-forward the identification and correction of roads. Various parts of the national capital have persistent complaints of potholes and delays in road maintenance. Officials say the system is expected to reduce response time for road repairs and improve monitoring of on-ground work.

Initially, the project will start on a relatively small scale—1,000 km—and will be expanded based on the outcome. Out of Delhi’s road network of over 18,000 lane kilometres, managed by agencies including the PWD, MCD, NDMC, DSIIDC, and I&FC, the PWD maintains around 1,386 km of key arterial roads.

As part of the project, a mobile application will enable periodic road inspections through geo-tagged video capture, with AI used to detect potholes and cracks and classify defects based on severity.

It will further support the automatic generation of location-based inspection reports, which will eventually enable easy review of defects identified on specific road stretches.
Another AI-based monitoring system will also be developed alongside for critical construction equipment, including rollers and pavers.

“The system will detect the operational status of the equipment, including working and idle conditions, using video analytics. This will help in assessing the utilisation of the covered equipment. In addition, the new system will support validation of compaction activity in the case of rollers. The output will help confirm whether compaction-related operations are being carried out,” the document, seen by News18, states.

In the case of pavers, the system will support validation of laying operations, and the output will help confirm whether paving-related operations are being performed.
The newly developed system will also generate utilisation and activity reports for the monitored equipment to support the review of equipment operation and field performance.

The project is to be completed in 12 months, including setting up an Integrated Command and Control Centre (ICCC) with video analytics support for at least 1,500 cameras.
In terms of budget, Rs 23 lakh has been allocated for the design, development, deployment, and maintenance of the IRMMS software system for data processing. An additional Rs 4.50 lakh has been allocated for 50 cameras and telematics units with SIMs, including installation and commissioning for road construction equipment for ongoing works.

An additional Rs 5 lakh has been allocated for creating, processing, and supplying baseline data collected through IRMMS, AI-based software for road deformities like potholes, cracks, and damaged footpaths/central verges, using a mobile application mounted on vehicles along the identified roads for a road length of 1,000 km.

However, bidders will have to place their quotations for the final price. The project will be awarded to the lowest bidder following the evaluation of quotations, after which work on the system is expected to begin.