Cops’ Artificial Intelligence to unclog Delhi

The Delhi Traffic Police has envisaged an ambitious plan called Intelligent Traffic Management System (ITMS) based on “Artificial Intelligence” to curb traffic congestion in the national Capital. The proposal has found favours with the Ministry of Home Affairs (MHA) and has been accepted in principle.

“The project will be implemented in three to four phases. The first phase will be implemented by January 2019 and will cater to areas of New Delhi, South and Central districts. Subsequently, the plan will be expanded to other areas of the city,” said Depender Pathak, Special Commissioner of Police (Traffic).

All system of ITMS will be on a single communication network using latest technology in communication e.g. 4G, fibre-optic network.  The communication network would support future expansion redundancy against failures and having cyber security system against cyber-threats.

The project will act as a public Interface for dissemination of information to the general public. Predictive modeling based on Artificial Intelligence (AI) to anticipate traffic congestion and take pre-emptive action.

It is the first time that the Delhi Traffic Police will be using the data collected through traffic detecting sensors, CCTV cameras, Automatic Number Plate Recognition (ANPR) camera systems, speed violation detection cameras. The data, collected, will be stored in a data cloud and will be used to predict traffic in a congested area.

One of the crucial features of the ITMS project is use of Adaptive Traffic Control Signals. These will assist in monitoring area and junction wise traffic optimisation. Adaptive Traffic Signals integrated with road based sensors as well as Traffic Control Centre for AI-controlled dynamic traffic management.

All these technological advancements will be monitored through a Centralised Monitoring Room. One Integrated Traffic command and control room with 5 regional centres and circle level nodes will be set up for effective and remote traffic management.

Pathak told The Pioneer, “Insights generated at the traffic control rooms via the input nodes like the cameras, traffic volume sensors, field functionaries will be made available to public on real time basis. The IP- based public addressed system will be earmarked to the traffic junctions which will make the motorists aware about the traffic situation ahead on the route.

The Variable Message Signage (VMS) will inform the motorists to make intelligent choices regarding traffic situations

Pathak said, “The Intelligent Traffic Management System will be implemented in four phases. The first phase will be implemented from January in 2019. We will use the data from CCTV cameras which will be saved in the cloud. This project will help the Delhi Traffic Police in getting an overall traffic view across the city. The traffic inspectors can further zoom into an area specific traffic situation. This will allow us to have better control and effectively manage the traffic in congested areas.”

The much anticipated Intelligent Traffic Management System has the capability to process and integrate information from multiple sources such as the intersection controllers, traffic light installations, sensors, CCTV cameras.

The project will also have a provision for automatically changing the switching instructions for traffic lights, managing malfunctions and maintaining traffic light installations and signs.

Alert generation mechanism which can enable faster and efficient decision support to manage traffic regulation and enforcement mechanism for the designated area

The ITMS project consists of multidimensional Infra Red based sensors for traffic classification and volume count, multidimensional colorless laser sensors to detect vehicles, Image pattern mapping based traffic volume sensing and ATCS control, video incident detection cameras for online traffic management and proactive action.

The project will cater to prosecution through artificial intelligence: E-challan system will be integrated with the Traffic Command and control centre, video incident detection system to compliment the prosecution system, backend database access to remote servers of e-Sarathi, Vahan, Court and Integration with National Crime Databases and CCTNS for stolen vehicle interception and suspect vehicles.

Some of the crucial features of the smart traffic project include optimisation of the traffic flows without changing the existing road network of the city, integrating the CCTV cameras with existing command centre as well as integrating them with the crime tracking applications like Crime and Criminals Tracking Network Systems (CCTNS), allowing automated traffic challans for various traffic violations, development of e-challan system and integration with any existing system, data collection for trend analysis and forecasting to support planning for traffic, parking Management and Guidance system.

The other features include sensor-based traffic volume count and intelligent, part of the Artificial Intelligence and Predictive approach is part of the Adaptive Traffic Control System (ATCS), image pattern analysis for classifying multiple vehicle types, effective data mining for Artificial Intelligence (AI) based historical pattern analysis for future traffic management, emergency system management for the Delhi Traffic and transport modes, state-of-the-art cyber security system with appropriate data disaster recovery mechanism, Speed Control Signals (SCS) for dynamic speed limit changes by ATCS to accommodate more traffic on available road-space, integration with on board vehicle information systems to give real time information to motorists.

For this, the information exchange with the Google Maps, Twitter, Facebook and other social media platforms for trigger based real-time information availability to the motorists by arranging and analysing key parameters at the traffic command and control centers.

Some of the key features of the Intelligent Management System (ITMS) are Congestion Management: Predicting queue length, optimize signal timings based on intelligent inputs through Geographical Information System (GIS) map based modeling, Spot speed detection to be 3D Radar based technology,  Lane and Speed violation Detection System: Speed measurement at vulnerable stretches and accident prone spots and appropriate non-intrusive sensor technology complemented by video cameras is being proposed.