Amir Ali Amini

A transition already under way

The automation of mobility is advancing along two distinct paths, which follow neither the same pace nor the same business logic.

The first is the steady upgrading of the private vehicle. Driver-assistance systems are becoming standard and more capable, gradually leading drivers to delegate a growing share of the driving task. This automation “by accumulation” asks the city for no permission: it arrives with the turnover of the vehicle fleet.

The second is the emergence of automated services. Robotaxi fleets are in commercial operation in a growing number of cities, and automated shuttles are beginning to support the frequency of transit routes. Here the vehicle changes in nature: from an object of ownership, it becomes a mobile asset whose use can be reassigned through the day — moving people, delivering goods, repositioning itself.

Together, these two dynamics put municipal administrations in an uncomfortable position: the capacities needed to govern this transition — legal, technical, organizational — take years to build, while the services themselves can appear in a matter of months. A city that waits for operators to knock on its door negotiates from a position of weakness.

The central issue: the usage model, not the technology

The potential benefits are real, and the first of them is road safety. A large share of collisions involves human error — distraction, fatigue, impairment. By addressing that root cause, automation offers serious potential to reduce deaths and serious injuries, whether the vehicles are shared or privately owned. To that add possible gains in network efficiency and, if dependence on parking declines, the freeing of considerable public space.

But the technology guarantees nothing: everything depends on which usage model takes hold. A rollout that reproduced the single-occupant, privately owned car would expose the city to well-documented systemic risks: growth in distance travelled — empty repositioning trips, a lower perceived cost of travel time — feeding congestion and pressure toward urban sprawl; erosion of ridership and of the funding base for structuring transit; and deepening dependence on private platforms to orchestrate mobility.

A through-line: the VMSOA approach

Faced with uncertainty of this scale, two symmetrical failures await administrations: paralysis — waiting indefinitely for clarity — and gadget-action — multiplying pilot projects with no link to any purpose. A cascading planning approach avoids both.

The VMSOA sequence links five levels, each framing the next: the vision(the city one wants), the mission (the role the administration takes on in the transition), the strategies (the postures and operating principles), the objectives (measurable, time-bound) and the actions (sequenced, with named owners). The value of the cascade is not bureaucratic: it makes the decision criteria explicit. When an operator proposes a project, the answer no longer depends on the mood of the moment, but on alignment with choices deliberately made upstream.

The five steps that follow work through that cascade.

Step 1 — Anchor the vision in existing commitments

First mistake to avoid: drafting a vision for automated mobility from a blank page. A city already holds democratically adopted commitments — its land-use and mobility plan, its climate plan, a Vision Zero–style road-safety approach, universal-accessibility commitments. The vision consists of stating the role automation must play in service of those existing objectives: safety, equity, restraint in resource use, quality of living environments.

A simple test follows from this — modal complementarity: any deployment should reinforce, not weaken, structuring transit and active modes. Stated at the vision stage, that criterion later becomes an operational filter for evaluating each project.

Step 2 — Map jurisdictions and stakeholders

Legal clarity is a prerequisite. In most jurisdictions, vehicle certification and technical safety fall to senior levels of government: a municipality will not regulate the technology itself. Far from a dead end, that fact clarifies the real field of play — and it is a large one. The city controls the public domain: the roadway, the curb, parking, occupancy permits, certain fiscal levers, and its own procurement.

The mapping then extends to the ecosystem: private operators, transit agencies, senior governments, research communities, citizens. The deliverable of this step is a clear-eyed inventory of what the city can decide alone, what it can influence, and what it will have to negotiate.

Step 3 — Choose a posture: proactive integration

Between laissez-faire and outright blocking, each end of the spectrum is a bet on a future no one controls. The posture of proactive integration avoids betting: it consists of getting ready for several futures at once. Three principles give it structure.

Favour robust actions. Some interventions create value under any deployment scenario — strengthening governance, improving safety, establishing data requirements. These “no-regret” actions form the foundation of the plan.

Plan by capabilities rather than predictions. Rather than optimizing for a single model, equip the administration with the skills, tools, and adaptable infrastructure needed to govern a range of use cases as they arise.

Institutionalize monitoring. Continuous tracking of technological and regulatory developments, with periodic re-prioritization checkpoints, turns uncertainty into a management input rather than a source of inaction.

Step 4 — Structure the intervention inventory

The possible interventions are best organized by the jurisdictional levers identified in Step 2. Two families emerge.

Direct municipal levers — what the city can act on under its own authority:

• Internal governance. A working group dedicated to automated mobility coordinates across departments, clarifies mandates, sustains strategic monitoring, and offers stakeholders a single point of contact.
• The municipal regulatory framework. An operating permit for services running on the public domain establishes the contractual relationship between the city and operators. To it attach dynamic curb management — designated pick-up and delivery zones, reservable or priced — and eco-fiscal levers that reward uses aligned with the vision: shared, off-peak, complementary to transit.
• Physical and digital infrastructure. Automated vehicles require no municipal connectivity to operate; but the ability to extend their perception beyond direct line of sight — through connected infrastructure, mobility-management systems, digital twins — delivers real safety and planning gains. A reference architecture keeps those investments coherent. And in physical design, priority for active and collective modes must be reaffirmed, not renegotiated.

Shared levers — what requires coordination with senior governments and the wider ecosystem:

• Data governance. Mobility data is a strategic asset. Sharing requirements must be set from the first permit issued: they feed measurable monitoring indicators and the optimization of traffic management. After the fact, the balance of power no longer allows it.
• Authorization and safety. Robust impact indicators, safety frameworks grounded in recognized standards, emergency-services readiness, and a phased rollout — pilots, then expansion conditional on results — make it possible to capitalize on the road-safety opportunity while keeping the risks of entry into service under control.
• Influence and social transition. Formalized intergovernmental collaboration, participation in the dialogue on legislative reform, cooperation with industry and research, citizen engagement, and support for the workforce whose occupations will be transformed.

Step 5 — From inventory to adaptive action plan

The intervention inventory is not a grocery list to be executed in full. It defines the capacities the city would gain from acquiring in order to be ready — legal, technical, organizational. Moving to an action plan means prioritizing the robust actions, sequencing by dependency, assigning owners, and attaching indicators to each workstream.

Implementation and periodic review fall naturally to the working group created in Step 4. An automated-vehicle integration plan is a living document: its quality is measured by its capacity to adjust, not by its thickness.

The classic pitfalls

Five traps recur with striking consistency in municipal efforts.

1. Betting on a single scenario : all-robotaxi and status-quo alike are bets, not strategies.
2. Treating the file as a technology project : it is first a question of urban planning, public finance, and equity.
3. Waiting for services to arrive before setting data requirements : once services are established, the negotiation turns asymmetric.
4. Underestimating local conditions : winter climate, street geometry, and snow-clearing operations are design constraints, not execution details.
5. Confusing monitoring with waiting : observing without building capacity is choosing to be acted upon.

---

Readiness is a capability, not a prediction

No one knows which usage model will prevail, or how fast the transition will unfold. But a ready city is recognizable by precise traits: a vision anchored in its existing commitments, a clear-eyed map of its jurisdictions, a deliberate posture, capacities under construction, and a plan that accepts revision. The cities that invest now in these foundations will shape the transition. The others will be shaped by it.