On April 2, Pune's roads dropped to walking speed. Here's what the data saw.

At 2 PM on April 2, Pune looked like any other Wednesday. By 4 PM, 207 road segments across the city were in severe congestion — 34 times the number recorded at the same hour the previous day. Average speeds on the Hadapsar–Ramtekdi corridor dropped to 2.4 km/h. That's not slow traffic. That's slower than walking.

This wasn't a single cause. Three independent events — record-breaking rainfall, a religious procession drawing 50,000 devotees, and ongoing metro construction — collided in a 6-hour window that brought Pune's road network to a near-standstill. The question isn't why it happened. The question is: what does the data actually show, and what can cities learn from it?

Three events, one afternoon

At 12 PM, the Pimpri-Chinchwad Traffic Police began road closures across Hinjawadi, Wakad, and Pimple Gurav for the annual Bagad Miravnuk — the procession of Gramdaivat Mhatoba Devachi Yatra. The procession, starting at 3 PM from Wakad Gavthan and moving toward Hinjawadi village, was expected to draw 40,000 to 50,000 devotees. Key arterials including Kaspate Corner to Wakadgav Chowk were shut until 10 PM, diverting thousands of vehicles onto parallel routes.

Simultaneously, Pune experienced what meteorologists later confirmed was a 130-year record for April rainfall. Shivajinagar recorded 65 mm in just two hours. In parts of Kharadi and Chandan Nagar, 25 mm fell in 25 minutes — faster than the storm drains could handle. Waterlogging was immediate and widespread, submerging vehicles in low-lying sections and reducing visibility to near-zero on several stretches.

The third factor was structural. Ongoing metro construction and road-widening projects across the Katraj–Kondhwa corridor — which already handles over 15,000 vehicles per hour — had narrowed lanes and disrupted natural drainage channels. In Kothrud, construction was actively underway on a drain that was already overflowing. The infrastructure designed to handle water was itself under construction.

What the data actually shows

TraffiCure monitors 933 road segments across Pune at 2-minute intervals. On April 1 — a normal Wednesday — the city's evening peak hit between 6 and 8 PM, with 47 severely congested segments at 6 PM and 126 at 7 PM. Average city-wide speed during the worst hour was 17.7 km/h. Unremarkable for a Tier-1 Indian city.

On April 2, the congestion signature was entirely different. It started three hours early — at 3 PM, not 6 PM — and was an order of magnitude worse. At 3 PM, 182 segments crossed the severe congestion threshold. By 4 PM, that number hit 207. At 5 PM, 116 segments were in extreme jam — delays exceeding 100% of typical travel time — and 34 segments had speeds below 5 km/h. The city didn't just slow down. It stopped.

The Hadapsar–Wanowrie corridor tells the starkest story. The stretch from Sinchan Nagar Junction to Suryamukhi Chowk — an arterial that typically runs at 28–30 km/h — dropped to 3.4 km/h at 6 PM, recording a delay of 997% against its baseline. That means a trip that normally takes 3 minutes took over 30. Gadital to Megacenter hit 3.0 km/h. Ramtekdi to Kalubai Junction crawled at 2.4 km/h with an 836% delay.

The timeline tells the story

Cross-referencing the traffic data with the news timeline reveals exactly how the three events cascaded. At 12 PM, road closures began for the procession — but traffic was still flowing. The data shows normal speeds through 1 PM. At 2 PM, the rain started. By 2:30 PM, waterlogging reports emerged from Kharadi and western suburbs. TraffiCure's data shows the first speed drops at 2 PM — delay percentages jumping from 6% to 10% city-wide.

At 3 PM, the procession began moving. The combination of closed arterials in Hinjawadi-Wakad and flooded roads in Kharadi-Chandannagar created a pincer effect — commuters had fewer routes available, and the routes that remained were waterlogged. Average city speed dropped from 22 km/h to 19.4 km/h in a single hour. In the Hadapsar corridor, the collapse was sharper — from 26 km/h at 2 PM to 16 km/h at 3 PM to 11.8 km/h at 4 PM.

The peak hit between 5 PM and 6 PM, when the regular evening commute piled onto already-gridlocked roads. The Hadapsar corridor bottomed out at 6.7 km/h average, with individual segments at 2–3 km/h. Recovery didn't begin until after 8 PM — once the procession ended and water levels receded. By 10 PM, most corridors were back to normal.

What this means for cities

Events like April 2 are not unpredictable. The procession was planned. The construction was ongoing. Only the rain was a surprise — and even that, in a city that floods every monsoon, is a matter of when, not if. The problem isn't that these events happen. It's that without full-network visibility, traffic operations teams can't see the cascading effects in real time.

A camera-based system monitoring 200 junctions would have seen congestion at those junctions. It would not have seen the 34 segments running below 5 km/h on service lanes and connector roads. It would not have detected the corridor-wide collapse from Gadital to Ramtekdi — because those mid-block segments have no cameras. And it would not have registered that the congestion started at 3 PM, not 6 PM — because the 15-minute refresh rate of traditional systems would have missed the rapid speed drop between 2:30 and 3:00 PM.

Software-only intelligence doesn't prevent rain. It doesn't reroute processions. But it sees the full picture — 933 segments, every 2 minutes — and that visibility is the difference between reacting to a gridlock at 5 PM and seeing it build at 2:30 PM.

The bottom line

April 2 was a stress test that Pune didn't schedule but its traffic network took anyway. Record rainfall, a 50,000-person procession, and construction on critical corridors — all at once, on the same afternoon. The data shows exactly when, where, and how the collapse happened: 207 segments in severe congestion, speeds at walking pace for six hours, and a recovery that didn't start until the procession ended and the water receded.

Cities that can see this in real time can respond to it. Cities that rely on camera feeds at 200 junctions will learn about it the next morning — from newspaper headlines, not from dashboards. The difference between the two is 933 monitored segments, a 2-minute refresh rate, and the willingness to look at every road, not just the ones with hardware on them.