Ubiquitous Connectivity: Sunderland’s Smart City Playbook

A smart city doesn’t fail because it lacks ideas. It fails because the basics aren’t there: reliable connectivity, trusted data sharing, and public services that work for everyone—not just the digitally confident.

That’s why Sunderland is a useful case study for anyone working on mākslīgais intelekts publiskajā sektorā un viedajās pilsētās. In a SmartCitiesWorld podcast conversation, Sunderland City Council’s smart city leadership and Sunderland University’s technical leadership describe a strategy that’s refreshingly practical: build the city’s “digital fabric” first, then use it to solve real problems in health, education, sustainability, and the local economy.

Here’s the stance I’ll take: AI in public sector and smart cities is only as good as the connectivity and governance underneath it. Sunderland’s story is about making that foundation ubiquitous, inclusive, and flexible—so AI can be deployed responsibly at scale.

Why “ubiquitous connectivity” is the non-negotiable layer

Ubiquitous connectivity is the operating system of a modern city. If you want AI-supported e-governance, real-time transport analytics, or sensor-driven sustainability, you need networks that are available where people live, work, learn, and access services.

In Sunderland’s roadmap, connectivity isn’t treated as a tech perk. It’s treated as a public service enabler. That framing matters because it changes the questions city leaders ask:

• Not “How fast is the network downtown?”
• But “Can residents access digital services across neighbourhoods, income levels, and life situations?”

The AI connection: better data, faster feedback loops

AI in smart cities lives on feedback loops: collect data → interpret it → decide → act → measure results. Without consistent connectivity, those loops break.

When the connectivity layer is strong, cities can support:

• AI-assisted operations (e.g., predicting asset failures, dispatch optimisation)
• Real-time situational awareness (e.g., congestion patterns, event planning)
• Digital inclusion programs that don’t collapse under peak usage

A practical rule I’ve found helpful: if a service can’t work reliably on a bad day (storms, events, demand spikes), it isn’t ready for AI automation. Sunderland’s emphasis on city-wide connectivity is a direct answer to that reality.

Inclusion isn’t a slogan—design it into the infrastructure

Digital inclusion has to be engineered, not announced. Sunderland’s smart city narrative consistently returns to “no one left behind,” and that’s more than a moral statement—it’s a performance requirement.

If your city rolls out AI-enabled public services (appointments, benefits, reporting issues, permits) but connectivity and access are uneven, you don’t just create inconvenience. You create two tiers of citizenship: those who can use the system and those who can’t.

What “inclusive AI” looks like in public services

Inclusive AI in the public sector is less about futuristic interfaces and more about dependable, boring wins:

1. Multiple channels by default: online + phone + in-person support
2. Identity and access designed for real life: shared devices, limited data plans, intermittent connectivity
3. Plain-language service design: fewer steps, clearer choices, error-tolerant forms
4. Bias-aware processes: human review paths, audit trails, and clear appeal routes

Sunderland University’s focus on equal access to education reinforces the same point: if connectivity is patchy, the benefits of digital systems concentrate among people who already have advantages.

Snippet-worthy take: A city can’t claim it has “AI-ready public services” if residents can’t reliably get online to use them.

Partnerships that actually serve the local community

Smart city partnerships work when the city stays in the driver’s seat. The podcast highlights Sunderland’s commitment to forming strategic partnerships—locally and globally—while maintaining a clear test: does this partnership genuinely benefit residents?

That’s the right filter, and it’s also a safeguard against a common smart city failure mode: vendor-led projects searching for problems.

A partnership checklist for AI and digital infrastructure

If you’re building an AI program in e-governance or smart city operations, partnerships should be evaluated like critical infrastructure—not like marketing collaborations. Here’s a checklist that maps well to Sunderland’s approach:

• Outcome clarity: What measurable public value will be delivered in 6–12 months?
• Data governance: Who owns the data, who can access it, and under what conditions?
• Interoperability: Can solutions integrate via APIs and common standards, or do they lock you in?
• Security model: How are identity, encryption, monitoring, and incident response handled?
• Skills transfer: Does the project build capability inside the municipality and local institutions?

The inclusion of a major university partner is especially strategic. Universities can provide neutral research capacity, testing environments, and talent pipelines—while helping the city avoid the trap of “black box” AI deployments.

Sustainability and the digital economy: use the same data twice

The best smart city data is reused across multiple goals. Sunderland’s “digital fabric” is described as supporting sustainability, the digital economy, and efficient city operations. That’s exactly how cities should think about it.

Instead of building isolated systems (one for environment, one for mobility, one for economic development), you build shared capabilities:

• Connectivity and sensor networks
• Data platforms and integration layers
• Analytics and AI models
• Service design patterns

Concrete examples of AI-enabled, connectivity-dependent wins

Even without naming specific Sunderland deployments, the model supports practical use cases that many Latvian municipalities and public agencies are actively prioritising:

• Smart infrastructure maintenance: AI models that flag abnormal water usage or predict streetlight failures (requires reliable telemetry)
• Traffic flow analysis: combining camera/sensor data and events calendars to adjust signalling and reduce congestion (requires stable, low-latency connectivity)
• Energy optimisation in public buildings: anomaly detection on heating/electricity loads, paired with automated work orders (requires building connectivity and integrated asset registers)
• Citizen reporting triage: AI that categorises issues (potholes, waste, lighting) and routes them to the right team with suggested priority (requires accessible digital channels)

The point isn’t “AI everywhere.” The point is AI where it reduces waste, time, and emissions, and where the city can prove it with data.

Flexibility beats perfect roadmaps

Sunderland’s emphasis on flexibility is a mature smart city move. Cities are complex systems; rigid roadmaps tend to fail because new regulations, funding cycles, emergencies, and political priorities reshape what’s feasible.

Flexibility doesn’t mean chaos. It means building modular capabilities that let you swap tools, add services, and scale what works.

A simple operating model for “real-world problem-solving”

If you want to mirror Sunderland’s pragmatic approach, structure initiatives around repeatable cycles:

1. Pick one pain point tied to a service outcome (e.g., missed waste collections, long wait times)
2. Confirm data availability (what you have, what’s missing, what’s sensitive)
3. Pilot in one area with clear success metrics
4. Operationalise (support model, training, procurement, maintenance)
5. Scale or stop based on evidence

This is where public sector AI programs often stumble: they run pilots, but they don’t plan the “boring middle” of operational ownership—monitoring, retraining, model drift, user support, and policy compliance.

Snippet-worthy take: A pilot is not a strategy. Operations are.

People also ask: what should a city do first?

Start with connectivity and governance, then target one service outcome. Sunderland’s story reinforces a sequence that works:

“Should we build a data platform before AI?”

Yes—at least a minimal one. AI needs clean(ish) data, shared definitions, and controlled access. A lightweight integration layer plus clear governance beats a giant platform nobody uses.

“How do we keep AI inclusive in e-governance?”

Design for mixed access: mobile-first forms, assisted digital support, and fallbacks when systems fail. Measure who is not using the service and why.

“What makes partnerships safe for the public sector?”

Contracts that guarantee data rights, interoperability, auditability, and an exit path. If you can’t leave a vendor without losing your operational capability, you don’t have a partnership—you have dependency.

Where this fits in the “Mākslīgais intelekts publiskajā sektorā un viedajās pilsētās” series

This series is about practical AI: e-pārvaldes improvements, infrastructure management, traffic analytics, and data-driven decision-making. Sunderland’s connectivity-first story is a reminder that these outcomes don’t start with algorithms. They start with a city-wide commitment to digital infrastructure, service inclusion, and partnerships that deliver local value.

If you’re planning your next step—whether that’s an AI-powered citizen service, a sensor rollout, or an analytics program—borrow Sunderland’s principle: build the digital fabric so public value can scale. Then prove it with one service at a time.

The forward-looking question that should guide your 2026 planning is simple: which city service would improve fastest if connectivity were truly ubiquitous—and what data would you need to make that improvement measurable?