Every day in 2023, close to 100 pedestrians died on Indian roads. The Ministry of Road Transport and Highways counted 35,221 pedestrian deaths that year, about one in five of the 1,72,890 people killed, according to an analysis of the MoRTH data by OpenCity.
A large share of that risk sits at signalised junctions, where cars, two-wheelers, cyclists and people on foot all compete for the same few seconds of green. A transportation engineering team at NICMAR University of Construction Studies, Hyderabad, has built an intersection safety system that tries to predict those conflicts before they happen.
The problem with the yellow light
Traffic engineers have a name for the most uncertain moment at a signal. The US Federal Highway Administration defines the dilemma zone as the stretch of road on an approach where drivers are unsure whether to stop or go when the light turns yellow.
Drivers caught in that zone either brake hard or speed through. FHWA research links that hesitation to rear-end, turning and sideswipe crashes.
Most dilemma zone models were built around vehicles alone. They look at speed, distance to the stop line and signal timing. On Indian streets, that leaves out a big part of the picture: pedestrians stepping off the kerb, cyclists filtering through traffic and crowds gathering at the corner.
How NICMAR’s intersection safety system works
The patented innovation, which its inventors call a Vulnerable Road User (VRU) Influence Modeling System, starts from a simple observation. A driver’s stop-or-go decision changes when people are near the crossing.
The intersection safety system adjusts the baseline probability that a vehicle is in the dilemma zone by factoring in what vulnerable road users are doing at that moment.
Sensing and tracking
The intersection safety system pulls data from several traffic sensors at once and time-synchronises it. From that combined feed, it estimates the paths of both vehicles and VRUs, such as pedestrians and cyclists, moving through the junction.
Measuring conflict in space and time
Next, the intersection safety system runs a spatiotemporal conflict assessment. In plain terms, it checks where each road user is likely to be in the next few seconds and flags where those paths may cross.
The influence model
The core of the patent is an influence-modelling framework. It takes the standard vehicle-only dilemma zone estimate and modifies it using VRU behaviour, crowd density, position and movement.
It also weighs vehicle dynamics, the geometry of the road and the signal timing plan. The output is a predictive risk picture that reflects how people and vehicles interact, rather than a fixed rule applied to every junction.
Why intersection safety matters now in India
The urgency is in the data. India recorded 4,80,583 road accidents in 2023, a 4.2% rise over 2022, according to the MoRTH report as covered by PTI. Two-wheeler riders made up 44.8% of those killed.
Pedestrians are the fastest-growing concern. OpenCity notes that pedestrian deaths rose by about 2,400 in a single year and that their share of fatalities has roughly doubled in eight years, from under 10% to more than 20%.
In some regions the picture is starker. Pedestrians accounted for 42.69% of road deaths in Delhi in 2023, per the same analysis.
Cities are already investing in adaptive signals, cameras and traffic command centres. Those systems generate the sensor data an intersection safety system needs, which gives a model like NICMAR’s a natural place to plug in.
Traction and proof
NICMAR’s work is at the research-to-application stage. The system is covered by a patent, according to the team, and is designed for proactive intersection safety management and for reducing vehicle-pedestrian and vehicle-cyclist conflicts.
The innovation comes out of NICMAR Hyderabad, a university campus at Shamirpet focused on construction, infrastructure, project management and clean energy. Its research arm runs a seed money scheme for faculty projects and a doctoral programme, which gives the work an institutional base for further testing.
The next proof point is a field pilot at a live signalised junction, where the model’s predictions can be compared against observed near-misses and crash records.
The researcher behind the model
The work is led by Dr. Suprabeet Datta, a transportation engineering researcher and Assistant Professor at NICMAR University of Construction Studies, Hyderabad.
His focus on vulnerable road users reflects how Indian traffic actually behaves. Mixed traffic, informal crossings and dense crowds at junctions are the norm, and most imported signal-safety models were never calibrated for them.
Building an intersection safety system inside a construction and infrastructure university also places it close to the engineers and planners who design intersections in the first place.
Why investors should watch this intersection safety system
Timing. Pedestrian deaths are rising faster than overall road deaths, and cities are spending on smart signal infrastructure. Both trends point to demand for tools that predict risk rather than count crashes after the fact.
The wedge. Most dilemma zone tools treat the intersection as a vehicle-only problem. NICMAR’s patent puts pedestrians and cyclists into the risk calculation, which suits Indian mixed traffic far better.
Business model logic. A patented, software-led model can travel through several routes: licensing to signal and traffic-management vendors, integration with city command centres, or a spin-out that sells intersection audits to road authorities. The team has kept the work bootstrapped so far.
What de-risks it. A paid pilot with a city traffic authority or an infrastructure contractor would be the milestone to watch. Validated results from one busy junction would make the case for scaling across a corridor.
Readers following how sensing and analytics are changing Indian infrastructure can also see how Indrones turns drone data into geospatial intelligence and how AgriCPS reads radar satellite data for farms. More on the institution is available on the NICMAR Hyderabad campus page.
The road ahead
India has spent decades widening roads for vehicles. The next phase of road safety will depend on how well junctions protect the people who walk and cycle through them.
NICMAR’s intersection safety system offers a model built for that reality, from a research team that designs for Indian streets as they are. If field pilots confirm what the patent proposes, every signalised junction in the country becomes a place where a few seconds of better prediction can save lives.
FAQ
What is an intersection safety system?
An intersection safety system uses sensors and models at a signalised junction to spot likely conflicts between vehicles, pedestrians and cyclists. NICMAR’s version predicts how vulnerable road users change a driver’s stop-or-go risk.
What is a dilemma zone at a traffic signal?
It is the stretch of road before a signal where drivers are unsure whether to stop or go when the light turns yellow. Hesitation there is linked to rear-end and turning crashes.
Who developed the NICMAR VRU influence model?
It was developed at NICMAR University of Construction Studies, Hyderabad, with work led by transportation engineering researcher Dr. Suprabeet Datta.