India
68 projects in 10 cities
The atlas documents 68 projects across 10 cities in India, the earliest dating to 1981: 4 with a documented success, 3 documented failures, 5 scaled back, 56 still running without a verdict. Every entry carries an evidence grade — the strongest here is peer-reviewed (4 entries) — and links its sources.
What the record shows
The oldest instrument in India's record is not a dashboard but a relay. Since 1981, Mumbai's grid has been able to cut itself loose, and Kolkata's private utility CESC runs the same manoeuvre. When the regional grids collapsed on 31 July 2012 — the largest blackout in history by population affected — CESC's island kept Kolkata running while every other district of West Bengal went dark. When lines tripped outside Mumbai on 12 October 2020, the island separated and then collapsed within minutes. What followed is as characteristic as the failure: India's power minister blamed human error, Maharashtra's energy minister backed a cyberattack theory, and the firm that had reported Chinese-linked intrusions into India's power sector called the connection unsubstantiated.
Most of what came after carries one federal stamp. The Smart Cities Mission, launched in June 2015, financed more than 8,000 projects across its 100 cities, worth around ₹1.6 lakh crore, with its mission period extended to 31 March 2025. Its signature product is the integrated command and control centre — operational, by the ministry's count, in all 100 cities. The atlas documents the genre: Ahmedabad's centre is demonstrably in use, yet its operator's own pages cannot agree whether the video wall carries 6,000, 7,000 or 7,200 cameras, and what is public about Pune's centre is largely its operator's own award submission. More entries here rest on the operator's own reporting than on any other source, and where operator and outside numbers meet, they can point in opposite directions: Bengaluru's traffic police report up to 25% shorter travel times on corridors re-signalled since 2024, while citywide rush-hour speed fell from 18 to 16.6 km/h over the same period.
The sharpest question this record asks is what a reading is allowed to do. Ahmedabad's heat action plan — written after a May 2010 heat wave left an estimated 1,344 excess deaths — bought almost no sensing: it converts a temperature forecast into numbered alert bands handed to a nodal officer with defined duties, and a pre/post evaluation found the all-cause death rate ratio on 47°C days fell from 2.34 before the plan to 1.25 after it — a design its authors call ecological and a pilot. The same city's air plan publishes its index and attaches no mandatory restriction to any band. Under GRAP, an air-quality threshold has had legal force since 2017 — construction bans, generator bans, curbs on older vehicles — yet CEEW's independent analysis found the stages invoked only after the threshold had already been crossed on 13 of 17 occasions in 2025. The counter-purchase stands in the same city: two 24-metre smog towers whose own pollution regulator told the National Green Tribunal that, on the measured 12–13% reduction within 110–200 metres, about 40,000 such towers would be needed to cut Delhi's pollution by 17% — after atmospheric scientists had argued, before either was built, that an open airshed cannot be filtered.
The record's most contested purchases watch people. Delhi's public works department has installed roughly 280,000 street cameras since 2018, and in March 2025 the government ordered a physical audit to check how many of the 263,000 recorded cameras exist — while Delhi Police treats an 80 percent similarity score on its facial recognition as a positive match, put the system's accuracy at 2 percent and then under 1 percent in its own court filings, and operates it under no specific law. Telangana's police centre in Hyderabad cost over ₹600 crore; Amnesty International's density mapping found 53.7% and 62.7% of two sampled neighbourhoods under camera coverage. When an instrument finds an enforcer, it tends to be a judge: Mumbai's dust-sensor mandate was widely ignored until the Bombay High Court ordered enforcement in January 2025 — and the BMC's December 2025 affidavit still counted 662 of roughly 1,954 active sites without a working monitor.
On rails, India buys the highest grade of automation and then walks toward it. Delhi Metro launched what India's government called the country's first driverless train operations in December 2020 — and removed the attendants from the Magenta Line only in stages between May 2024 and May 2025. Kolkata's Green Line has been trial-running automatic operation since August 2025, with formal clearance still pending after the safety commissioner's July 2026 inspection — the infrastructure and the certification are different projects on different timelines. Mumbai's Aqua Line, India's first fully underground metro, carried under 20,000 riders a day against a 400,000 forecast over its first 136 partly-open days; Bengaluru's Yellow Line stood structurally complete from late 2023 while its Chinese-ordered trains sat behind a border clash, opening in August 2025 with three trainsets of fifteen — the Takshashila Institution put the cost of the delay at 1,866 crore rupees against the 410 crore saved by taking the cheapest bid.
What survives, again and again, is the unglamorous telemetry, while the contract on top of it dies. Janmarg in Ahmedabad discontinued the smartcard behind its first-in-India fare claim at the end of 2025 while the operator's older website still listed it as the way to pay; what has actually lasted sixteen years is a locator on every bus reporting to one room. Chennai's sensor-and-app parking survived neither contractor turnover nor its successor's overcharging scandal; Delhi's doorstep delivery — about 430,000 service requests handled — ended not by decision but because the helpline contract ran out. Delhi's transit feeds supply Google Maps and over 50 organisations, yet reporting in 2026 found GPS units offline, stale route files and not one bus stop showing live arrivals. The counter-model is Namma Yatri, built in consultation with Bengaluru's auto-rickshaw drivers' union: zero commission, code open-sourced under AGPL-3.0, and — on the operator's own figures — 100 million rides with about 1,600 crore rupees paid directly to more than 600,000 drivers by June 2025. If India's entries agree on one lesson, it is the one Ahmedabad's heat plan states plainly: the instrument is the cheap part. What separates the projects that changed something from the ones that only watched is an addressee — a named officer for a forecast, a legal threshold for a reading, a safety commissioner for a train, a judge where nothing else answers.
Sources beyond the atlas
Facts above that do not come from a linked entry rest on these:
- PIB / Ministry of Housing & Urban Affairs, 3 July 2024 — Smart Cities Mission extended till March 2025 — mission since June 2015, 100 cities selected by competition, 8,000+ projects worth around ₹1.6 lakh crore, mission period extended to 31 March 2025
- PIB press note — 10 Years of Smart Cities Mission — launched 25 June 2015; integrated command and control centres operational in all 100 smart cities (as of 31 March 2025)
The shape of India
Computed from the 68 entries below — a profile, not a score.
Cities
All 68 projects
✔ Every published entry is checked against the public sources it cites before publication.
Show 56 more projects in IndiaShow fewer
Deeper reads
Where India appears in the atlas's own analysis.
📎 Cite this country record
1001 Smart Cities (2026). “Smart-city projects in India.” The Smart City Atlas. https://1001smartcities.org/countries/india/ (last verified 2026-08-27). Data: CC BY 4.0.
The underlying data is free to reuse with attribution — see the open data page.
