Urban Lifeline Waterlogging Monitoring & Flood Dispatch

🤖 AI-generated summary of the sources below · 🔎 Last verified 2026-08-06

This overview was generated with AI from the public sources listed below and checked against them before publication. It has not had a complete human fact-check — treat it as a starting point and follow the sources.

How this atlas is made · What “last verified” means

Nanjing's water facilities centre built a waterlogging platform as the second phase of its smart-water programme, part of Jiangsu's urban lifeline safety pilot. It added 544 IoT devices — cameras, electronic water gauges, flood alarms, level and flow meters, smart manhole covers, vehicle video — and pulled in 1,293 sensors already run by phase one, districts and utilities. More than 6,000 river outfalls in the main urban area are tagged with a live three-colour submersion state; 134 residential compounds and road flood points are instrumented; 79 flood-prone spots use gauges that upload a depth reading and a photograph, which the system screens by AI instead of a person watching video.

Live 📅 since 2024 🤖 Uses AI

Outcome🔄 Ongoing

Evidence🏛️ Official

Run byNanjing Municipal Water Facility Management Centre (南京市水务设施管理中心), under the Nanjing Water Bureau

Scale🏘️ District

📊 Impact

Across the rain events of 5–6 and 19–20 August 2024 the platform pushed 12 flood warnings, 18 incident work orders, 8 video-AI alerts and 26 dispatch instructions, and linked 31 complaints from the 12345 hotline and 127 police reports. On 20 August a 6:41 complaint about standing water reached the named responsible officer as a work order four minutes later. That meiyu season delivered 461.8 mm citywide against a normal 284.5 mm.

🎓 Lesson

The gain here is not the sensor, it is what the sensor replaced: a rule of thumb. Pre-storm river drawdown used to be set from pump forebay design levels, and at one station the design range ran from 3.0 m to 4.5 m — so operators simply pulled everything to the bottom. Outfall telemetry gives them a reason to choose a number. The reach is still small: the sample zone is 5.8 km² of a 6,587 km² municipality.

🌱 Environment Flood resilience Water Sensors & IoT Artificial intelligence

Sources

↗ Nanjing Municipal Government / Nanjing Daily (official)13 Sep 2024, the single source for almost every figure here. 『项目累计建设摄像头、电子水尺、积水报警器、液位计、流量计、智能井盖、车载视频等物联感知设备共544套，同时共享了智慧水务一期和各区、各行业单位的传感设备1293套』; the 5.8 km² sample zone; 『主城地区沿河道分布的6000多个排口全有标记，并以红色、黄色、蓝色三色显示』; the Changxiang pump station 4.5 m/3.0 m design range and the old blanket drawdown; 『我市主城六区134个居民小区、道路易积淹水点已加装积水报警器…』; 『这样的电子水尺已覆盖南京79个易积淹水点』; the August 2024 counts and the four-minute work order. Operator self-report — no independent evaluation exists.

↗ Xinhua Daily (Jiangsu, state media)Parallel report of the same completion announcement, used only to confirm the project is a Jiangsu urban-lifeline pilot and not a vendor claim.

Continue exploring

Compare this project

🇨🇳 Automated Waterlogging Monitoring & Early-Warning System · Tianjin Two Chinese cities instrumenting flood points under the same provincial lifeline push, both reporting their own dispatch counts. 🇹🇭 Bangkok Flood Control Centre: Drainage Telemetry & Street-Level Forecasting · Bangkok Telemetry into a staffed room in both; Nanjing screens the gauge photographs by model rather than having someone watch the video. 🇨🇳 Flood & Typhoon Command System · Shanghai The Yangtze delta's two flood command systems, one organised around typhoons at the coast, the other around meiyu rain inland.

Explore the wider context

📟 A Reading Is Not a Decision 🌊 Flood resilience guide 🏙️ Nanjing city profile 🌱 Environment projects 📖 Adaptive traffic signals 📖 Flood early-warning system 📖 Non-revenue water (NRW)

📎 Cite this project

1001 Smart Cities (2026). “Urban Lifeline Waterlogging Monitoring & Flood Dispatch” — Nanjing, China. The Smart City Atlas. https://1001smartcities.org/projects/nanjing-waterlogging-lifeline/ (last verified 2026-08-06). Data: CC BY 4.0.

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

🚩 Report an error