How do you measure weight? Kilograms.
Length? Meters or centimetres.
And distance? Well, that depends on where you live.
Ask someone in Delhi, Mumbai or Bengaluru how far their office is from home.
“About an hour.”
Not 15 kilometres. Not 20 kilometres. An hour. Strange, isn’t it?
Somewhere along the way, our cities seem to have invented their own unit of distance: time.
And worse, that time keeps changing. The same commute may take 30 minutes one day and 90 the next. So we leave early, check alternative routes and still wonder whether we’ll arrive on time.
According to 2025 Traffic Index, Delhi ranked fourth among Indian cities in congestion levels, with drivers losing an estimated 104 hours annually to rush-hour congestion. That is more than four days spent waiting in traffic.
Of course, Delhi has more vehicles than its roads can comfortably accommodate. But not every traffic jam requires another flyover. Sometimes, the problem is how we manage the roads and junctions we already have.
On 30 September 2026, the Union Cabinet approved a ₹1,789.52-crore Intelligent Traffic Management System (ITMS) for Delhi.
The plan is substantial. It includes AI-powered adaptive signals at 1,092 locations, automated enforcement at 1,029 locations, emergency-vehicle priority and coordinated traffic management across 42 major corridors.
To start with the basics, let’s understand the difference between Conventional signals and Adaptive signals. Conventional signals generally follow preset timings, while adaptive signals use cameras and sensors, connected and powered by Artificial Intelligence to adjust timings according to real-time traffic conditions. Neighbouring junctions can also coordinate, creating a green wave of fewer unnecessary stops.
Additionally, automated enforcement will also detect violations such as speeding, red-light jumping etc, enabling digital notices and challans. The entire system will operate through integrated traffic-control centers.
Sounds promising, doesn’t it?
This week, Policy Mandala explores whether Delhi’s new traffic system can make journeys faster and more predictable, and what it will take for smarter signals to improve mobility across the city.
But anyone who has driven regularly in Delhi knows that atleast some signals already respond to traffic conditions, and automatic challans do land up on our doorsteps and emails. In fact, this idea has been in the making for over two decades.
First enabled in 2005, Delhi introduced area traffic control at 46 junctions. This technology was to be expanded ahead of the 2010 Commonwealth Games, when the Union government approved a ₹200-crore traffic and communication network scheme, covering approximately 87 roads and 210 kilometres in its first phase. However, the plans failed to materialise adequately.
Automated cameras followed suit in 2019, and in 2022, C-DAC (leading research lab by central government) began preparing a detailed project report for ITMS, with 1,146 junctions and blinkers surveyed out of 1,471 identified.
And Delhi isn’t alone in this.
Bengaluru has adaptive signals at 169 junctions, while Mumbai reports 258 adaptive signals. In London, a similar system called SCOOT has delivered reported reductions of around 12% in traffic delays at upgraded signals, while our South Asian friend, Singapore, uses coordinated traffic signals across its network.
So, this is not new. Neither for Delhi, nor for India, nor for the world. There is significant experience to build upon. However, what changes now is the proposed scale and integration of Delhi’s system.
And there is a reason to be optimistic about this:
An October 2026 Delhi Traffic Police simulation projected that traffic delays could fall by approximately 40% on stretches including Nelson Mandela Road, 10% on Ring Road and even 2% on the GT Road–Karnal Road stretch.
These are projections, not achieved improvements. But they suggest that intelligent signals could make a considerable difference on some roads.
Why such differences within the roads?
Well, simply because congestion has different causes.
One junction suffers from poor signal timings; another from excess vehicles. Elsewhere, parking, encroachments or badly designed turns obstruct movement. Clearing one junction may simply shift queues elsewhere.
The selected 42 corridors for this project include Ring Road, Outer Ring Road, Mehrauli–Gurgaon Road, Nelson Mandela Road, the DND Flyway–Ghazipur Road stretch, amongst others. However, without knowing what share of Delhi’s daily traffic uses these corridors, their citywide impact remains difficult to estimate.
And this brings us to the complexity of Delhi’s traffic problem.
According to Delhi’s Economic Survey 2025–26, the city’s registered vehicle stock increased by approximately 6.44 lakh vehicles, or 7.93%, in 2025–26 alone, reaching 87.61 lakh.
Meanwhile, the Survey lists approximately 18,500 lane-kilometers of roads across the agencies it reports, as of March 2025. Expanding road capacity is difficult in a city where land is already committed to buildings, settlements and other infrastructure. The available figures do not establish a directly comparable annual road-growth rate.
And as with most problems in public policy, there are no silver bullets here.
To understand Delhi’s traffic, consider three interconnected factors:
Traffic demand: How many vehicles and people travel, their commuting patterns and the availability of public transport.
Road infrastructure: The roads, junctions, parking facilities and usable space available to accommodate movement.
Traffic management: How efficiently signals, enforcement and incident-response systems manage the available capacity.
The challenge becomes more complicated because responsibilities are divided.
Delhi Police, under the Union Ministry of Home Affairs, regulates traffic. PWD, MCD, NDMC, NHAI and other agencies maintain different parts of the road infrastructure. This makes coordination central to building better infrastructure. And after all of this, traffic demand depends partly on how people like you and me choose to travel.
It's important to realise that the adaptive signal can use available space better, but one cannot create space where none exists. Yet we believe that the opportunity with ITMS is much bigger than traffic lights.
Let's consider three possibilities.
First, ITMS can improve real-time traffic management, coordinating signals, identifying accidents and informing commuters about congestion, diversions and available parking.
Second, it can help identify recurring bottlenecks. If a junction jams every evening, traffic data may reveal illegal parking, excessive turning traffic or poor road design, allowing engineers to address the underlying problem. A fixed signal leverages one time data, an adaptive signal builds recurring data for future.
Third, it could contribute to long-term public transportation planning. You see, adaptive signals note down data on recurring public movements, and it helps cities create better public transportation infrastructure.
In August 2026, Delhi’s PWD announced an AI-assisted traffic study covering 450 kilometers of roads to inform future infrastructure and transport decisions. ITMS could complement such exercises through continuous traffic data.
The same information could help identify where buses need priority or where an ambulance requires coordinated green signals.
After all, moving 100 vehicles is not necessarily the same as moving 100 people.
But traffic data becomes valuable when it leads to coordinated action.
And coordination is precisely where Delhi faces challenges.
In 2025–26, authorities identified 62 congestion hotspots and proposed 215 interventions. By June 2026, however, 56 had been dropped or deemed infeasible, often because of land constraints, narrow roads or ongoing projects.
Identifying the problem was easier than fixing it.
An intelligent control room might identify an obstruction instantly, but another agency may need to remove it.
Delhi needs to connect institutions as effectively as it connects signals.
Enforcement presents another challenge.
In February 2026, Delhi Traffic Police had 4,901 personnel against 6,102 sanctioned posts, leaving roughly one in five positions vacant. Automated monitoring could help extend their reach with even limited capacity.
But detection alone does not ensure compliance.
Delhi issued approximately 74.5 lakh challans and notices in 2024. Reporting in February 2025 found around 94% remained unpaid at that point.
Effective enforcement will require accurate evidence, accessible appeals and timely resolution. Otherwise, more cameras may simply mean more unresolved notices. We hope the government brings some action on that front as well.
Coming back to the ITMS, the plan involves implementation in three phases over 24 months, followed by five years of operations and maintenance.
The investment indicates a substantial policy commitment. The real test will be implementation.
Will travel times improve? Will buses become more reliable? Will ambulances move faster? And will information from traffic-control centers help other agencies fix recurring problems?
Independent assessments across upgraded and yet-to-be-upgraded corridors could help answer these questions.
Finally, let’s return to where we started.
Will ITMS change how we measure distance? In kilometers or minutes?
Perhaps it doesn’t matter. Here, the unit matters less than the number in front of it.
A journey may take 30 minutes or an hour. What matters is knowing, with reasonable confidence, how long it will take. That predictability allows people to plan their work, spend time with their families and organize their day.
Delhi’s traffic challenge is enormous, and solutions will require sustained effort across several institutions.
If ITMS can help make our journeys more predictable, it will have achieved something that matters to every commuter, regardless of whether they measure distance in kilometers or hours.
Co-Authored by: Kumar Subham & Avdhesh Pathak
