Pune
Ujjawal.Gayakwad, CC BY-SA 4.0, via Wikimedia Commons
🇮🇳

Pune

Country
India
Population
3.1M
Projects
8
First project
2015
Best evidence
Official
Last verified
2 August 2026

🤖 AI-generated city profile, summarising the source-checked entries below — how this works

Pune's smart-city record is largely a record of how each piece was paid for, and of what happened after the money landed. A twelve-year concession bought roughly 80,000 LED streetlights with no upfront cost to the city; a municipal bond, the first Indian civic issue in fourteen years, part-funded a round-the-clock metered water programme; the adaptive signal network needed a second payment to its contractor in 2024, on top of installation and maintenance, before its performance claims held up. The financing arrived intact each time and the follow-through lagged. The water programme was at about 75 per cent physical progress eight years in. The new LED fixtures became a standing target for theft off the city's own streets. And the construction-dust mandate put sensors on every large site and a public dashboard behind them before anyone published a penalty for a site that fails.

The atlas documents 8 projects in Pune, the earliest dating to 2015: 8 still running without a verdict. Every entry carries an evidence grade — the strongest here is official (4 entries) — and links its sources.

Start here

Three entries that show what Pune is actually doing.

The shape of Pune

Computed from the 8 entries below — a profile, not a score.

Outcome ongoing 8
Evidence official 4media 4

How Pune got smart

Pune began by instrumenting things it already ran. Its transport company had a command centre monitoring routes, schedules and breakdowns through GPS on every vehicle from the 2015 bus-corridor expansion, feeding what became an all-green fleet — around 400 electric buses inside a fleet that by 2022 ran on compressed gas and electricity with no diesel bus left, unusual among large Indian operators. September 2016 brought two projects in the same month. A twelve-year design-build-finance-operate-maintain-transfer concession swapped roughly 80,000 conventional street lights for LEDs monitored through about 1,900 feeder-level SCADA units at no upfront cost to the city, and the property tax department started geo-tagging the city, recording owners, photographs and building footprints and flagging construction put up after a December 2015 cutoff.

Then came the two things the city had to finance itself. The round-the-clock metered water programme started in 2017, part-funded by a ten-year municipal bond — the first Indian civic issue in fourteen years — toward reservoirs, bulk and smart metering and distribution works, with the city divided into 327 district metered areas so that inflow and outflow can be compared and leakage and theft located. In March 2020 the smart-city company opened an integrated command and control centre on Sinhagad Road, pulling ambulance tracking, patient heat-mapping, traffic, street lighting, waste-vehicle and utility feeds into one room. Its best-documented moment is the pandemic, when it coordinated fifteen ward-level operations rooms and, by its own account, logged and resolved 4,982 citizen requests. Its newest job, from 2026, is mapping roughly 2,000 kilometres of buried water, drainage and cable lines so the same street is not dug up twice.

2022 added two networks that automate in different senses of the word. Signals at 125 junctions came under an adaptive traffic management system adjusting green time to live density and reporting to the command centre — and the rollout was rocky enough that glitches and complaints forced repeated review meetings between the corporation, the police and the smart-city company, and a further payment to the contractor in 2024 to fix performance, before the city's cited figures settled at roughly 15 per cent less congestion, about 10 per cent faster speeds and close to seven in ten vehicles clearing junctions without stopping. It still degrades in the monsoon. On the rail side, both metro lines were fitted with communications-based train control under a 2019 contract covering 31 trains, completed across the network by late 2024: continuous automatic protection and operation, with a driver aboard for doors and monitoring. That is supervised automation, not the driverless capability the technology theoretically allows.

The most recent addition is a mandate with a gap in it. Following a state pollution-board directive, a corporation circular in December 2025 required every construction project of 5,000 square metres or more, and every corporation infrastructure project, to fit sensor-based particulate monitors with on-site display boards; a real-time dashboard went live in February 2026, integrating the site feeds so that high-pollution sites can be flagged for inspection, after a co-location study with the Indian Institute of Tropical Meteorology validated the sensors against reference instruments. At launch the corporation had published no penalty schedule for a non-compliant site, and it is not confirmed that the dashboard is visible to anyone outside the municipality. Around it, the older programmes were still finishing: the metered water scheme at about 75 per cent physical progress in September 2025, its tank and pipeline component at 85 per cent in February 2026, both well past their original dates, while a separate drone survey meant to find unregistered properties collapsed when the surveying agency's fee, set at 30 per cent of recovered revenue, proved unworkable.

All 8 projects in Pune

On a time axis, by the year each one started — 2015 to 2025. Pick a dot to read what it was.

✔ Every published entry is checked against the public sources it cites before publication.

How the pieces connect

Editorial links between Pune's own entries — the city as a system, not a list.

Show 1 more connectionsShow fewer

What's next

Plans and announced trajectories — not yet documented results.

Pune has been unusually inventive about paying for infrastructure and unusually quiet about what the payment bought, and each of the open questions is a measurement the city could publish without buying anything. The property cadastre is a decade old and no revenue-recovery figure has appeared, which is the whole difference between a data project and a demonstrated one. The bond-funded water programme has to finish before anyone can ask whether 327 metered areas actually reduced losses. The signal network has the city's own numbers and no independent before-and-after, on a system that already needed a second contractor payment to reach them. The clearest test of all is the newest: a sensor mandate that flags a site for inspection is only real once inspections follow, and this dashboard went live before any published penalty for failing it.

Deeper reads

Where Pune appears in the atlas's own analysis.

Closest comparisons

Two cities the atlas reads next to Pune — the reason first, then the entries it rests on. Chosen by hand from the Pune threads below, not counted out of them.

  1. 🇮🇳 Kolkata

    Two Indian cities that signed the same two long contracts: LED street lighting with a control layer on a multi-year concession, and a CBTC metro that has stayed supervised since its first signalled section opened.

  2. 🇮🇳 Hyderabad

    Two Indian cities running the same three systems — adaptive signals, a parcel map and a metro that could drive itself — and differing in what each is aimed at: Pune's map collects tax, Hyderabad's reclaims public land.

Show 22 further city threadsShow fewer

Every editorial link from Pune's entries into another city, strongest first. Each one is a statement about two projects, not about two cities — the three above are the ones the atlas is willing to make at city level.

Official strategy and sources

↗ Pune Municipal CorporationThe organization driving the city's smart-city work.
📎 Cite this city

1001 Smart Cities (2026). “Smart-city projects in Pune, India.” The Smart City Atlas. https://1001smartcities.org/cities/pune/ (last verified 2026-08-02). Data: CC BY 4.0.

The underlying data is free to reuse with attribution — see the open data page.