Smart Cities in Practice: Successes, Failures and Lessons

"Smart city" has become a catch-all term for using data, sensors and digital technology to run cities better: moving traffic more efficiently, saving energy, improving public services and making government more responsive. Some projects have delivered real improvements. Others have been expensive disappointments or raised serious concerns about privacy and who controls urban data. A handful of real examples, both successes and failures, show what works and what to watch for.

Singapore: pricing roads by demand

Singapore introduced Electronic Road Pricing in 1998, charging drivers automatically when they pass gantries on congested roads, with prices that vary by location and time of day. Combined with strict limits on car ownership and heavy investment in public transport, it has helped keep traffic moving in one of the world's densest cities. The city-state is moving to a satellite-based system. Singapore also uses detailed digital models of the city for planning, from flood management to building placement.

Lesson: technology works best when combined with clear policy goals, in this case reducing congestion, and good alternatives such as public transport.

London: congestion and data

London's congestion charge, introduced in 2003, uses cameras and automatic number plate recognition to charge vehicles entering central London. Transport for London has also opened much of its transit data to developers, which allowed apps such as Citymapper to offer live journey planning. Opening public data in this way has become a model for other cities.

Lesson: sharing data openly can create benefits the city itself would never have built.

Los Angeles: synchronized traffic lights

Los Angeles completed the synchronization of all of its roughly 4,500 traffic signals in 2013, using sensors embedded in roads and a central control system that adjusts timing to traffic conditions. City officials reported modest improvements in travel times on major routes.

Lesson: smart traffic systems can help, but they can't eliminate congestion caused by sheer numbers of cars.

Copenhagen and Barcelona: designing for people

Copenhagen has used sensors and "green wave" signal timing on some routes so cyclists riding at a steady speed hit successive green lights, supporting its very high rates of cycling. Barcelona has used sensors for smart irrigation, parking and waste collection, but some of its best-known changes, the superblocks, are low-tech: rerouting traffic around groups of blocks to free streets for pedestrians and green space.

Lesson: the smartest solutions are not always the most technological. Data can guide design, but street design often matters most.

Songdo: a city built from scratch

Songdo, in South Korea, was planned from the ground up in the 2000s as a model smart city, with sensors throughout, pneumatic waste collection and extensive green space. It has attracted residents and businesses, but it filled more slowly than planned, and critics have described parts of it as quiet and lacking the street life of organically grown cities.

Lesson: technology does not create a community on its own.

Toronto: a project that was cancelled

In 2017, Sidewalk Labs, a sister company of Google, proposed a sensor-filled neighborhood on Toronto's waterfront called Quayside. The project faced sustained criticism over data collection, privacy and the role of a private tech company in governing public space. Sidewalk Labs withdrew in 2020, citing economic uncertainty during the pandemic. Toronto later moved forward with a different plan for the site.

Lesson: public trust, transparency and clear rules on data are essential, and residents expect a say.

Amsterdam and Helsinki: open platforms

Amsterdam has run a long-running smart city program that brings together the city government, companies, universities and residents to test projects such as smart energy meters, shared mobility and data on canal boat traffic. Helsinki is known for opening up city data and for early experiments with mobility as a service, combining public transport, bike sharing and taxis in a single app. Both approaches emphasize partnerships and open data rather than a single large contract.

Lesson: small, open pilots that can be tested and scaled are often more effective than grand plans.

Water, energy and lighting

Some of the most practical smart city projects are invisible to most residents. Many cities have replaced streetlights with LED lamps that can be dimmed remotely or brightened when sensors detect movement, cutting energy use. Water utilities use sensors and smart meters to find leaks in pipe networks faster, reducing waste. Smart electricity meters let utilities balance demand and help households see when they use the most power.

Concerns to watch

Privacy and surveillance: cameras, license plate readers and facial recognition can track people's movements. Several cities have restricted police use of facial recognition.
Security: connected infrastructure can be vulnerable to cyberattacks, as ransomware attacks on city governments have shown.
Vendor lock-in: cities can become dependent on a single company's proprietary systems.
Equity: services that rely on smartphones and apps can leave out residents without them.
Cost: sensors and systems need ongoing maintenance and replacement, not just upfront investment.

What makes a city genuinely smart

The most successful examples start with a specific problem, such as congestion, energy waste or slow permits, rather than with technology. They involve residents, set clear rules on data, measure results and keep human judgment in charge. Many of the most valuable improvements are unglamorous: reliable buses with live arrival times, online services that save residents a trip to city hall, and well-maintained infrastructure.

For residents, it is worth knowing what data your city collects and how it is used. Many cities publish open data portals, privacy policies for sensors and cameras, and opportunities to comment on new projects at public meetings.