🇹🇭

Thailand

10 projects in 1 city

🤖 AI-generated country overview of source-checked atlas entries — how this works

The atlas documents 10 projects across 1 city in Thailand, the earliest dating to 2007: 1 scaled back, 9 still running without a verdict. Every entry carries an evidence grade — the strongest here is peer-reviewed (1 entry) — and links its sources.

What the record shows

Bangkok is unusually willing to publish the number that embarrasses it, and unusually unwilling to license it. Almost the entire Thai record dates from a single administration that made self-documentation its method — the citywide problem-reporting service launched days after the 2022 governorship changed hands, and the tree register, the data catalogue and the policy scorecard all followed within weeks. What that produced is a record you can argue with, which is rarer than it sounds. The recurring gap in it is not between promise and delivery. It is between detecting and collecting, between publishing and licensing, between warning and capacity.

The million-tree register is the model. Every planting is logged as an object — district crews, companies and residents photograph a specimen, pick the species and drop a map pin, residents through a messaging-app account — and the entries appear on a live public map. The register then does something a campaign number almost never does: it exposes what the headline contains. When the city declared its first million reached in July 2024, only 35 per cent of the records were trees; the rest were shrubs and climbers. It also shows who did the work, because of that first million, district offices logged 804,791 records and residents' communities 2,281. A per-object register makes a political number auditable, and this one convicts its own press release. No survival audit has been published, which is the obvious next question and the one still open.

The enforcement entries are where detection and collection come apart hardest. The city's camera fleet is run by the traffic department rather than the police and passed 65,015 units at 16,623 sites — but most are standalone recorders rather than pooled feeds, a 2017 count found 38,404 of 53,249 cameras unconnected, and about a fifth of footage requests came back empty. Where the AI layer works it works well and then stops: cameras at a hundred sites logged 90,907 footpath-riding violations in the year to September 2025, and 10,977 fines were paid — roughly one in eight. The same ratio governs the air-quality lever. Bangkok publishes readings from 523 monitoring points and colours every district against a national index tightened in June 2023, and red carries real consequences: on 24 January 2025 the city closed 352 of its 437 schools and, borrowing disaster-law powers, barred unregistered trucks from nine inner districts. Its low-emission-zone pilot kept 404 trucks a day out and measured particulates 15.6 per cent below surrounding areas. Enforcement was thin in the same way — 1,547 recorded violations produced 28 fines — and the threshold bites only where the city holds the legal lever: it can close its own schools outright, it can only ask employers to send staff home, and it cannot reach the burning upwind, which a peer-reviewed model traces to roughly 30 per cent of the region's haze-season fine particulates originating outside Thailand.

Instrumentation buying warning time rather than capacity is the drainage story exactly. The flood control centre runs two weather radars, a rain-gauge network and canal, river and road level sensors into a room staffed around the clock, with a forecasting system that predicts street-level flood depth at four-metre resolution and, since 2024, three-hour rain nowcasts. Crews still work the 200 pumping stations, 243 gates and four tunnels themselves, and pumping pre-emptively on a forecast is still being studied. On 3 November 2025, 131.5 mm fell in Lak Si against a drainage system designed for 60 mm, and the roads went under anyway. Bangkok can see a storm three hours out and still cannot pump faster than the rain falls.

Where the obstacle is a contract rather than a sensor, the pattern holds too. Bangkok's thirteen rail lines sit under three agencies and their concessionaires, and the common card promised from 2016 launched on two lines in 2018 before being relaunched years later as a bank-issued card. Open-loop contactless carried the integration instead — bank cards have worked on two MRT lines since January 2022 and became mandatory there in June 2026 — while the BTS lines still take only their own card. A network-wide flat fare was budgeted and opened for registration in August 2025, slipped past its October launch awaiting legislation, and was dropped by the cabinet in June 2026 in favour of a capped band from 2027. Open payment standards fixed the turnstile and not the concession contracts underneath, which is where integration actually stalled. Electrification went the other way and moved faster: a private concessionaire took over routes the debt-laden state operator gave up, and by April 2026 had 2,273 licensed electric buses on 124 routes — of which 1,576 ran on a typical day, on the operator's own tap-in card that rail does not accept, with its supplier's then-chief executive accused of fraud by the securities regulator in 2024. Handing the flagship green fleet to a concessionaire tied it to one company's balance sheet, and to its supplier's.

Which brings the record back to its own transparency. Open Bangkok opened a draft city budget for public comment, runs a catalogue of more than 900 datasets from 98 agencies, and tracks every campaign policy down to project and budget line. The scorecard is self-graded — the governor's count of policies completed against policies promised was stated four days before an election he won — and the catalogue is downloadable rather than licensed, so the group that visualised the budget had to request the files from city offices. Publishing a scorecard is the cheap half of transparency; licensing the data so outsiders can recompute it is the half Bangkok has not done. The reporting service has the same shape: more than 1.2 million reports received and around 81 per cent resolved on the city's own figures, with councillors warning that tickets get closed to protect departmental scores and reports needing agencies outside the city administration stalling. A closed ticket records that an office responded, not that the problem stayed fixed. If the entries below agree on one lesson, it is that Bangkok has built an unusually honest set of instruments and has consistently stopped one step short of letting anyone else grade them.

The shape of Thailand

Computed from the 10 entries below — a profile, not a score.

Outcome ongoing 9scaled back 1
Evidence peer-reviewed 1official 9

Cities

All 10 projects

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

Deeper reads

Where Thailand appears in the atlas's own analysis.

📎 Cite this country record

1001 Smart Cities (2026). “Smart-city projects in Thailand.” The Smart City Atlas. https://1001smartcities.org/countries/thailand/ (last verified 2026-08-26). Data: CC BY 4.0.

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