Smart Cities Aren’t Smart If They Don’t Work for Everyone

A city is not an app

The phrase “smart city” usually arrives accompanied by a glossy rendering: autonomous shuttles gliding past trees, sensor-equipped bins, and a person staring happily at a phone as though they have just discovered electricity. It is all very tidy. Real cities, as those of us who have tried to cross one in the rain know, are less cooperative.

I am an engineer, so I do appreciate a well-placed sensor. I have spent enough time designing temperature-monitoring and IoT systems to know that useful data can prevent failures, save stock, reduce waste and make people’s working lives easier. Technology is not the problem. The problem is treating technology as the city, rather than as one component of a city that has to work for actual humans.

As a disabled person, I experience the gap between a polished digital promise and physical reality rather often. A council can publish live bus information, which is genuinely useful, but it is of limited comfort if the bus stop has no seating, the pavement is broken, the display is unreadable in bright sunlight, or the accessible route to the stop has been blocked by a badly parked e-scooter. The app may be intelligent. The journey is not.

Accessibility cannot be a later software update

Too many smart-city projects begin with the technology: cameras, connectivity, dashboards, predictive models and a procurement brochure featuring suspiciously cheerful pedestrians. Accessibility is then added near the end, often as a compliance exercise. This is backwards.

Disabled people should be involved before requirements are written, not invited to test a nearly finished system and politely identify the bits that exclude us. That means people with mobility impairments, sensory impairments, cognitive differences, chronic illness and fluctuating energy levels. “One disabled user” is not a representative sample any more than one temperature sensor tells you what is happening throughout a warehouse.

Consider a pedestrian-crossing app that lets someone request more time to cross. Good idea. But does it work without a smartphone? Is it usable with a screen reader? Can a person use it when their hands shake, their battery is flat, or mobile signal has chosen that particular street for an afternoon off? Does the crossing also provide audible and tactile signals? If not, we have built a clever exception mechanism around a basic piece of infrastructure that should already work well.

The least glamorous answer is often the best one: longer crossing times where needed, level surfaces, reliable dropped kerbs, clear signage, benches, toilets and staff who can help. Nobody will make a keynote slide about a properly maintained dropped kerb. It remains considerably more useful than an AI-powered kerb that sends a monthly report explaining why nobody crossed it.

Data is useful, but it is not neutral

Smart cities run on data, and data can reveal problems that have long been ignored. It can show where buses run late, where air quality is poor, which lifts fail repeatedly, and where footfall drops because a street has become inaccessible. Used carefully, that is valuable evidence.

But data reflects what we choose to measure. If a city measures vehicle flow in extraordinary detail while recording nothing about broken pavements, inaccessible bus stops or out-of-service lifts, it is making a policy choice disguised as technical objectivity. The dashboard will look impressively precise while quietly telling us that cars count more than people.

There are privacy issues too. Disabled people are often asked to share more information to receive basic support: location, travel patterns, health-related needs and identity details. Cities must collect the minimum necessary data, explain clearly why it is needed, protect it properly, and offer non-digital alternatives. Consent is not meaningful when declining means losing access to transport or a public service.

I am not anti-data. I am anti the casual assumption that gathering more of it is automatically progress. A sensor is not a moral compass, inconveniently.

Reliability is an accessibility feature

In embedded systems, reliability is never exciting until it fails. Nobody applauds a temperature alarm for working during a power cut, right up until the moment it prevents thousands of pounds of stock becoming an expensive soup. Public technology is much the same.

A smart city service needs graceful failure modes. If the app crashes, there should be a phone number. If the ticketing gate fails, staff should be empowered to help rather than recite a policy written by someone who has never been stranded. If a lift is out of service, the information must be accurate, immediate and available before someone reaches the station platform and discovers it through personal inconvenience.

This matters particularly for disabled people because a small failure can have a large consequence. A non-working lift is not a minor delay if it leaves you unable to leave a station. An unavailable accessible toilet can end a trip. A route change without clear notice can turn a routine journey into a logistical puzzle with fewer pieces and no table.

Designing for reliability means speaking to the people who live with the consequences, then funding maintenance. It is not glamorous, but neither is being stuck at the bottom of a staircase while a touchscreen confidently reports that your journey is “on track”.

Measure what makes life better

The best test of a smart city is not the number of connected devices, gigabytes collected or AI pilots announced. It is whether people can get where they need to go, use public services independently, breathe cleaner air, feel safe and participate in civic life.

Ask practical questions. Did accessible journey times improve? Are faults fixed faster in poorer neighbourhoods as well as wealthy ones? Can residents report problems in ways that suit them? Have digital services reduced bureaucracy rather than moved it onto people’s phones? Are disabled residents helping to decide what gets built and how success is measured?

A genuinely smart city is modest enough to start with the basics and wise enough to listen. Give me a reliable bus, a clear route, a working lift and a human being available when the system goes wrong. Add good technology where it helps. That is not a futuristic fantasy. It is simply a city behaving as though everyone who lives in it matters.