This direction closes the section with the same finding that appears in the mobility article, from a different angle, which is itself the point.
The pilot that never became a service
A decade of smart city investment has produced a large number of pilots and a much smaller number of durable services.
The usual explanations are that the technology was immature or that the city lacked digital skills. Both are sometimes true and neither is the main pattern.
Most smart city pilots worked. They ended because nobody owned them once the funding round closed.
Two structural reasons.
Pilots are funded as projects, services need budgets. A grant-funded pilot has capital and a fixed period. Turning it into a service requires ongoing operating budget, a team, a support arrangement and a place in someone's departmental responsibilities. None of that is in the pilot.
Success is measured as delivery, not adoption. The pilot is reported as complete when the technology is installed, which means nobody is accountable for whether anything changed afterwards.
What the direction covers
The scope: cities as systems, transport, housing, utilities, public space and connected infrastructure.
Four areas.
Planning and land use. How development is guided, what gets built where, and the statutory system that governs it.
Urban transport. Network design and the relationship between land use and travel demand, which connects to the mobility direction.
Infrastructure and utilities. Water, energy, waste and digital, and the coordination between them.
Urban data and digital. Sensing, modelling and the governance covered below.
The four governance blockers
Named explicitly, because a workforce plan has to staff against them.
Procurement. Public procurement is designed for accountability and for buying defined things. It handles a specification for 200 bus shelters well and an evolving digital service poorly. Framing the requirement, running a compliant process and structuring a contract that permits iteration is a specialist skill, and where it is absent projects either stall or buy something inflexible.
Data ownership. When a vendor operates a system, who owns the data it generates, who can access it, and what happens at the end of the contract. Cities that did not settle this at the outset have found themselves unable to switch supplier without losing their own historical data.
Departmental boundaries. Transport, planning, waste, housing and environmental health hold separate budgets, systems and objectives. A project whose benefits land in one department and whose costs land in another has no natural sponsor. This is the single most common reason cross-cutting initiatives die quietly.
Political cycles. Programmes span electoral terms, and priorities change. Building something that survives a change of administration requires it to be embedded in operations rather than associated with a person.
The capability that addresses all four is not technical. It is people who understand public sector operations, procurement and politics well enough to design around them, and who also understand what the technology can do.
Where this sits in the domain
Urban planning and smart cities is the sixth and final direction in Astra Trainer's engineering and built world domain, and the one that connects the other five to the city scale.
It pairs directly with future mobility and transportation systems in the space and mobility domain, where the same integration failures appear in a transport frame, and with AI, data and computing for the data platform and cybersecurity layers. Partners here are usually local authorities, agencies or consultancies. Lessons are five minutes, which suits officers with no protected development time. You can see the six directions here.
Planning capacity is the quiet constraint
An issue that is discussed as a housing problem and is substantially a workforce problem.
Planning departments in many jurisdictions have reduced capacity over a long period while the complexity of what they assess has increased: environmental requirements, viability assessment, design quality, heritage, flood risk and carbon.
Four consequences.
Determination times extend, which delays delivery of housing and infrastructure regardless of how much anyone wants them built.
Experienced planners are stretched thin, which reduces the capacity to negotiate better outcomes rather than simply processing applications.
Specialist input is scarce. Ecology, heritage, urban design and viability specialists are drawn on by multiple authorities, and the queue is the constraint.
Enforcement is under-resourced, which weakens the credibility of conditions that were negotiated.
This is a straightforward workforce issue with a straightforward framing: the planning system's throughput is limited by trained people, and treating it as a policy problem alone will not change the throughput.
Data about people is a different problem
Worth separating clearly, because conflating it with asset data is how cities lose public consent.
Data about infrastructure, such as how a bridge is performing or where a water leak is, raises few concerns. Data about people, including movement, behaviour and imagery, raises real and legitimate ones.
Four points that belong in any programme.
Legal obligations apply. Data protection law in most jurisdictions governs personal data, including data that can be re-identified, and public authorities frequently have additional duties.
Aggregation and anonymisation are not automatic protections. Movement data in particular is difficult to anonymise robustly, because patterns of travel are highly identifying.
Consent is mostly unavailable. People walking through a public space cannot meaningfully consent, which means the justification has to rest on other lawful grounds and on proportionality.
Public trust is the operating licence. Several high-profile urban data projects have been withdrawn after public opposition, not because they were unlawful but because the public did not accept them. Transparency about what is collected, why, and who can see it is a practical requirement rather than an ethical flourish.
The workforce implication is that information governance capability belongs in urban data teams from the start, not as a review at the end.
The roles, named
Planning officers and development management staff. The capacity constraint.
Urban designers.
Transport planners, shared with the mobility direction.
Urban data analysts and modellers.
Digital and smart city programme managers. The role that has to hold technology, procurement and politics together.
Procurement specialists with digital capability. Genuinely scarce and genuinely gating.
Information governance officers.
Infrastructure coordination officers, managing utilities and street works, which is unglamorous and directly affects daily life.
Who can be trained into it
Local government officers from any department. They understand how the institution works, where the budgets sit and what will and will not get approved, which is the knowledge these projects most lack and cannot buy.
Planning support and technical staff. Into planning officer roles, which is a direct intervention on the capacity constraint.
GIS and data analysts. Into urban data, needing the planning and governance context.
Transport and highways staff. Into broader urban systems work.
Procurement officers. Into digital procurement, which is a specific and learnable extension of what they do.
Community engagement and communications staff. Into consultation and consent-building, which determines whether schemes survive contact with residents.
Statutory process and data protection. Planning is a statutory function with legal procedures, consultation requirements and rights of appeal, and decisions carry legal weight. Processing personal data is governed by data protection law with specific obligations for public authorities, and surveillance technologies may carry additional requirements. Training builds professional understanding, and planning practice in many jurisdictions also requires professional membership or qualification. It does not confer any statutory authority, delegated decision-making power, or lawful basis for processing personal data.
What to take from this
Most smart city pilots worked and ended because nobody owned them after the funding closed. That is a governance outcome, not a technology one.
Procurement, data ownership, departmental boundaries and political cycles are the four blockers, and the capability that addresses them is institutional rather than technical.
Planning department capacity limits housing and infrastructure delivery and is rarely framed as the workforce problem it is.
Data about people carries obligations that asset data does not, movement data is hard to anonymise, and public trust is the operating licence.
And local government officers hold the institutional knowledge these projects depend on. They are usually consulted after the technology has been chosen, which is the wrong order.
Why do smart city projects fail?
Procurement designed for buying defined things rather than evolving services, unresolved data ownership, departmental boundaries where benefits and costs land in different places, and political cycles. The technology usually works.
Why do pilots not become services?
Because pilots are funded as capital projects with a fixed period, and becoming a service requires ongoing operating budget, a team and a departmental owner. Success is also measured as installation rather than adoption.
Is planning capacity a workforce issue?
Yes. Departments have reduced capacity while assessment complexity increased, which extends determination times, stretches experienced staff and under-resources enforcement. Throughput is limited by trained people.
What is different about data on people?
Legal obligations apply, aggregation and anonymisation are weaker protections than assumed, particularly for movement data, consent is mostly unavailable in public space, and public trust is the operating licence.
Who converts into these roles?
Local government officers from any department, who understand how the institution works; planning support staff into officer roles; procurement officers into digital procurement; and GIS analysts into urban data.