Chapter 04

What Is at Stake

How value can be created and destroyed over time

4.1What is at stake in each iODD domain

Connected autonomous mobility does not create new risks or new opportunities in isolation. It amplifies outcomes that are already latent in the way cities are governed, funded, and spatially organized. Where governance is coherent – where cities have built the sensing, rules, economic instruments, and design interventions described in the chapters that follow – CAM accelerates value creation across several levers at once. Where governance is absent or fragmented, it accelerates value destruction with equal efficiency. That asymmetry is a strong argument for acting early.

To make this concrete, consider a city the size of London. The governed upside – the value that structured deployment can create and that unstructured deployment will leave uncaptured or actively destroy – is conservatively in the range of $2.5 to $5 billion (£2–4 billion) annually. Each component is grounded in specific precedent and mechanism, not projection. London loses an estimated $6.5 billion (£5.1 billion) per year to traffic congestion: lost productive time, late deliveries, increased fuel costs, the operational inefficiency of public transport in congested conditions. Stockholm’s congestion pricing cut traffic 20 to 25 percent^EV35 (Chapter 6). Applying a comparable reduction to London’s congestion losses – achievable through a combination of autonomous fleet pooling mandates, routing governance, and pricing – would recover between between $1.3 and $1.6 billion (£1–1.3 billion) annually.

Property value capture along CAM-served corridors represents a second source of conditional value. The Elizabeth Line (Chapter 7), shows how reliably that uplift tracks distance from the station. CAM corridors can use routing, curb design, depot siting, and service obligations coordinated by the city to govern uplift.

The operating cost advantages of coordinated fleet electrification add a third dimension. Electric fleet operations cost 60 to 70 percent less per kilometer in energy terms than diesel equivalents, with total cost of ownership advantages of 40 to 50 percent when maintenance savings are included. Operators whose depot charging is coordinated with grid constraints through managed charging programs capture the full benefit of this advantage; operators forced to charge at peak tariff due to unmanaged load profiles capture a fraction of it. Cities that have established managed charging standards and utility coordination frameworks can direct some of these savings into lower fares, wider coverage, and longer operating hours that reinforce transit use rather than undermine it.

Table 4.1 Governed dividend sensitivity: annual value range by city type and governance posture. Values in US Dollars. Estimates are indicative, based on congestion reduction, property capture, electrification savings, and fiscal replacement mechanisms.

City archetype	Early governance (2026–27)	Delayed governance (2028–29)	No governance
Global city (8m+) e.g. New York, London	$2.5–5bn/yr; fiscal stability; competitive corridor premiums 6–12%	$0.6–2bn/yr; partial capture; fiscal erosion under way; retrofit costs double	Net fiscal loss $1.3–2.5bn/yr; transit under stress; unmanaged congestion; political backlash likely
Large regional (1–4m) e.g. Chicago, Toronto, Manchester	$250–750m/yr; depot coordination; transit agency-operator integration; managed grid load	$60–250m/yr; fragmented deployment; operator-led siting; grid stress emerging	Revenue loss $125–500m/yr; stranded parking assets; service gaps in lower-income areas
Mid-size (250k–1m) e.g. Portland, Austin, Bristol	$40–125m/yr; pilot-led learning; conversion-ready design; first-mover corridor advantage	$12–50m/yr; reactive adaptation; missed corridor advantage; grid bottlenecks	Net negative; parking revenue collapse without replacement; operator-defined terms

Value is conditional on governance. Cities and operators are on the same side of this asymmetry. The same governance investments that prevent value destruction enable value creation. The same mechanisms that protect city interests are the conditions under which operators sustain commercial deployment over the long run. The operators that have engaged most actively with governance – Waymo under California’s mandatory incident reporting regime, the major micromobility operators under Austin’s licensing regime built on the Mobility Data Specification (MDS) – have so far generally sustained more durable commercial trajectories than the operators that have not. What follows sets out what is at stake across five levers of value creation, and the coordination mechanisms by which the balance is struck in each.

4.2The five levers of value creation

The conditional dividend works through five levers: fiscal, energy, real estate, spatial allocation, and legitimacy. Each lever has a measurable value creation potential that proactive coordination can manifest, a corresponding value at risk that absent coordination accumulates, and a specific instrument through which the balance is struck. Chapters 5 through 8 build out the mechanisms; this chapter shows, domain by domain, what they are for.

Table 4.2 Five levers of value creation: the value at risk in each, the value coordination creates, and the instrument that determines which way the asymmetry resolves. Cross-references to the toolkit chapters that build out each instrument in detail.

Domain	Value at Risk	Value Created	Instrument (→ Ch)
Fiscal	Parking + enforcement + farebox cascade; s	Transaction- and deployment-fee structure scaling automatically with operator activity; predictable operator cost base; Portland’s micromobility framework as worked precedent	Fiscal model choice built into licensing conditions (→ Ch.6, sequenced Ch.8)
Energy	Concentrated capital cost	Fleet-as-battery and managed charging as distributed energy resource; operator energy savings captured at full benefit; grid headroom protected; Austin Energy as muni-utility template	Managed-charging standards in licensing + joint transport-energy planning with DNOs (→ Ch.5, Ch.6)
Real Estate	US: 50–70% non-residential land in surface parking^EV19. Dense markets: scarce operational land contested with housing, logistics, datacenters, urban industrial	US: conversion-ready commercial stock (Austin parking-garage reuse template). Dense markets: land-assembly with explicit trade-off frameworks; distributed-depot structure priced into asset values	Conversion-ready design specs + zoning protections + explicit trade-off frameworks for operational land (→ Ch.7)
Spatial Allocation	Curb capture; event-mode chaos ; VRU interaction failure	Dynamic curb pricing; event-mode governance flexibility; VRU-interaction operating conditions protecting cycling/pedestrian infrastructure	Operating-condition flexibility in licensing + spatial design patterns (→ Ch.6, Ch.7)
Legitimacy	Backlash + whiplash ; labor displacement at scale without transition mechanisms	Collaboration premium (illustrated – not proven – by Waymo’s continued SF operation vs Cruise’s October 2023 suspension); sustained operator confidence; durable commercial trajectories; workforce transition funded as licensing condition	Mandatory transparency + community benefit + workforce transition (→ Ch.6, Ch.8)

Fiscal

The fiscal opportunity in CAM is architectural: replacing eroding revenue streams with a framework designed around the externalities CAM creates. Transaction- and deployment-based fees – a fixed deployment charge per authorized vehicle, a per-trip charge that scales with utilization, and where applicable a curb-event charge – track the activity that produces the externalities the city is governing. Applied to CAM, the framework scales linearly with operator activity: when fleet deployment doubles, the revenue doubles. Portland has run this for micromobility since 2019 (Chapter 6). Who bears each fee – operator margin, rider fare, or slower deployment – is unknown. No jurisdiction has published an incidence estimate for a fee of this kind. Cities can explicitly design a fee schedule to be revised against evidence.

Without that structure, the value at risk is substantial. Parking revenue is the most visible exposure: English councils generated over $1.3 billion (£1 billion) in net parking surplus across 2023 and 2024 combined, with Westminster alone producing $96 million (£75.9 million) annually.^EV15 New York issued 16.1 million parking violations in fiscal year 2024, generating $1.088 billion in fine revenue; Chicago’s parking and traffic fines represent c. 7 percent of the entire operating budget.^EV15 Transit farebox compounds the exposure: TfL’s 58 percent farebox recovery ratio^EV17 – unusually high for a Western city – reflects pricing and service choices that autonomous fleets offering lower door-to-door journey times at undercutting price points could materially erode. No measured effect of robotaxi service on transit ridership has been recorded anywhere yet. The farebox exposure is prospective. In San Francisco, TNC competition accounted for roughly half of Muni’s ridership decline in the late 2010s, before CAM had entered the picture. Seventy-one percent of large US transit agencies are projected to face fiscal cliff risk within five years, before autonomous mobility adds a structural competitor to the farebox equation.^EV17 The exposure is a cascade across revenue streams governed by different agencies on different timescales.

The instrument is a fee structure written into licensing conditions, with corridor value capture reserved for the cases where its logic holds (Chapter 6; sequenced in Chapter 8). Cities that build the new revenue base before the old one collapses have headroom to invest in the rest of the stack.

Energy

The risk associated with energy concentrates in two forms. The first is capital cost: a 100-vehicle fleet depot in London is estimated to cost approximately $29 million (£23 million) in total, with roughly $295,000 (£232,000) attributable to each vehicle’s share of charging infrastructure, grid connection, and civil works.^EV37 Depots represent a significant investment in specific urban locations, and will therefore compete directly with other infrastructure priorities.

Figure 4.2 The energy lever in one picture: a depot's 24-hour charging load, unmanaged. The evening peak breaks through the local grid ceiling; managed charging moves the same energy under it.

The second is operational risk: in unmanaged deployments, 65-75% of fleet charging demand concentrates in a four-hour evening window^EV26, creating peak loads that distribution infrastructure was not designed to absorb. This stress triggers curtailment (brownouts), damages grid assets, exposes operators to tariff penalties or exposes end users to rate hikes. At 20% EV penetration in a dense urban area, modeled studies project 8 to 14% increases in evening peak network load^EV26. Cities without a coordination framework between transport licensing and energy infrastructure planning will discover the consequences only after a depot is built, when increasing transmission capacity will cost far more.

Managed charging turns a stability threat into a powerful tool for increasing efficiency. A 500-vehicle electric fleet whose charging is shaped by a managed-charging program contributes capacity to the local distribution network rather than draining it. The same depot infrastructure that creates the energy value at risk under unmanaged conditions becomes a distributed energy resource under managed ones. Demand management programs improve operator economics, too, enabling them to avoid peak rate charges.

Austin Energy’s demand-side programs include time-of-use incentivized charging and solar power purchasing. These, and many others, serve as an operational template. A city-owned utility can negotiate charging schedules with depot operators in ways that benefit both the grid and the fleet. Cities without a municipal utility cannot simply replicate this. An investor-owned utility aligns with its regulator, not with the city’s strategic plan, so the equivalent lever runs through regulatory interventions rather than direct coordination. These may include rate cases, connection-policy consultations, hosting-capacity disclosure requirements – the mechanisms are different in kind.

Managed-charging standards should be written into licensing conditions (Chapter 6), backed by the load-profile sensing in Chapter 5. The institutional model is a joint transport-energy planning between the city, the transport authority, and the distribution network operator. Such cross-agency bodies do not yet exist in most regions. Building it is the highest-priority Power-domain investment in the 12-month governance horizon.

Real estate

Real estate splits by market type, and the framework has to hold both. In low-density US cities – and to a lesser extent in some Australian, Canadian, and Gulf cities – the opportunity is conversion. The parking land that defines the US value at risk also represents the largest single source of conversion-ready commercial property in many cities. Where the city owns the underlying asset, the opportunity is immediate: Austin’s parking-garage reuse work – city-owned garages re-purposed as charging hubs with ground-floor retail and last-mile logistics tenancy – is the template at municipal scale. The city captures the revenue, the operator absorbs the conversion capital, and the asset transitions to its next use without the long political contest that surrounds private-sector parking de-listings. In the distributed-depot cities of Chapter 3 the opportunity is structured land-assembly: explicit trade-off frameworks between CAM depots, bus electrification facilities, last-mile logistics yards, and housing, supported by zoning protections for distributed depot networks. In both regimes, the opportunity is to price land allocations under uncertainty into asset values and policy decisions rather than leave them latent.

In US cities, a large share of non-residential urban land – in some downtowns more than half – is dedicated to parking. US parking revenue fell by an average of 5.9 percent annually between 2017 and 2022^EV19, driven by reduced commuter travel, transit substitution, and the early effects of TNC competition. Investors and asset owners who have not begun planning conversion pathways are accumulating stranded asset exposure on a timeline current valuations do not reflect. In land-constrained markets, there is no parking glut to convert; the binding constraint is the scarcity of operational land where CAM depots, charging facilities, and logistics yards must compete directly with other uses. The exposure is not parking-revenue erosion but the opportunity cost of land allocations that may be irreversible. Both regimes are governed by the same principle – land allocation under uncertainty creates path dependency that valuations do not yet price – but the binding constraints, and therefore the governance instruments, are different in each.

The instrument is a conversion-ready design specification embedded in planning policy (Chapter 7), with zoning protections and explicit trade-off rules for operational land in constrained markets (Chapter 6). The conversion-ready premium, detailed in Chapter 7, is cheap insurance against stranded-asset risk.

Spatial allocation

The spatial allocation opportunity is the active governance of how shared public space is allocated, in real time, between competing uses. Dynamic curb pricing turns the curb into a managed resource that responds to demand; New York’s Smart Curbs and Washington DC’s curbFlow^EV32 (Chapter 6). Event-mode flexibility, defined in Chapter 3, converts festival, emergency, and weather-driven peaks into managed operating conditions. Structured operating conditions for vulnerable road user interaction – measured, audited, and enforced through licensing – protect the public realm investment in cycling and pedestrian infrastructure that cities have spent two decades building. The unifying mechanism in all three is operational authority: the city’s right to set, monitor, and adjust the operating conditions of fleet activity in shared public space.

Without operational authority, the exposure across spatial allocation is the accumulation of conflicts the city has limited instruments to resolve. Curb capture is the steady-state mechanism: fleet operators optimize pickup and drop-off for operational efficiency, not curb equity, and the consequence in New York is double-parked delivery vehicles blocking at least one travel lane for about 25 percent of the day on monitored Brooklyn streets.^EV24 Event-mode chaos is the discrete-event mechanism – the Austin festival clusters described in Chapter 3. Vulnerable road user exposure is the most consequential: cyclists and pedestrians moving in patterns autonomous sensing has not resolved, in road space whose protections do not carry over into fleet operating conditions. In all three, the city holds responsibility for outcomes without the authority to manage them.

The instrument is operating-condition flexibility written into the fleet license (Chapter 6), made implementable by the spatial design patterns in Chapter 7. Operating conditions belong in the license from the start, not as a fallback when problems emerge.

Legitimacy

Legitimacy entails sustained public trust and aligned operator behavior. Formal governance frameworks facilitate voluntary collaboration that ungoverned environments do not. These include transparent operating conditions, incident reporting protocols, community benefit agreements, workforce transition mechanisms. Operators that understand the rules of the road and know which residents the city is protecting are materially more willing to invest in the partnerships, capital commitments, and information-sharing arrangements that produce the city’s preferred outcomes.

Austin Energy explicitly protected resident interests, while remaining open to operator partnerships. It was transparent about what counts as success. This example, and countless others, undermine the conventional “government versus innovation” framing.

The risk is spiraling from initial deployment friction to an outright prohibition. Paris banned shared e-scooters by referendum, with 89 percent of voters supporting the ban on a turnout of roughly 7.5 percent^EV10 – a result showing how little mobilized opposition it takes to end a mode when there is no channel for grievance, and one that reflected accumulated frustration with a deployment managed without democratic accountability rather than specific hostility to the vehicles.

Organized opposition to Waymo is already active in San Francisco, where vehicles have been blocked, had their tires slashed, and in one case been set alight; in Boston it has taken the form of union and legislative campaigns. The route from friction to prohibition is short when governance is absent, and legitimacy once lost is the hardest constraint to recover from.

Autonomous mobility will displace professional drivers (taxi, private hire, bus, delivery and logistics) who represent a large, organized, and politically visible workforce. Transport for London licenses over 100,000 private hire vehicle drivers; the New York Taxi and Limousine Commission oversees more than 200,000 for-hire drivers; one recent estimate puts 4.5 million US jobs and $168 billion in wages at risk from autonomous vehicle adoption^EV3230, with the losses concentrated among immigrant, non-graduate, and lower-income workers for whom driving was accessible work. Concentrated job loss without transition mechanisms is a legitimacy crisis the technology will not have time to absorb.

The instrument is a clear transparency protocol, community benefit, and workforce transition, written into the license as conditions (Chapter 6). Workforce transition mechanisms are named here as a license condition. The Urban Autonomy Lab’s research agenda includes detailed design of such programs in the future, potentially including, for example, operator contributions to transition funds, scaled to displacement and paid over the licensing period.

The governance investment is small relative to the legitimacy exposure it protects against. The political return on that investment compounds over the deployment window.

4.3The collaborative balance

Across all five levers the outcome depends on coordination, and no single actor can produce it alone: the fee structure needs the city, the transport authority, and the operators; managed charging and depot siting need the DNO and the regulator as well; conversion-ready standards need planners, asset owners, and developers; and legitimacy needs the workforce and the affected communities at the table.

In each case the coordination work – building the institutional model that allows these stakeholders to converge on workable arrangements – is the governance investment that carries everything else. The technical instruments described in the chapters that follow are the operational expression of that coordination work. They do not produce the balance on their own; they are the mechanisms through which the coordination is enacted.

The alternative posture causes value destruction across every area. It is also the posture that produces the outcomes operators themselves want least, such as the Paris referendum and the various Tempe-style incidents that set deployment back by years. The operators that have engaged most actively with governance have so far also been those with the most durable commercial deployment. The cities that have invested most consistently in coordination capacity have attracted the largest commercial commitments. The asset owners that have specified conversion-ready design early have captured the optionality premium. The pattern is consistent across all five levers.

None of this is a claim about goodwill. This is how CAM will get done. The city needs the operator’s data, the operator needs the city’s instruments, the DNO needs the transport authority to treat energy as a siting question, and asset owners need planners to make conversion-readiness a condition of consent.

The next four chapters describe the how: sensing (Chapter 5), rules and economic instruments (Chapter 6), design and investment (Chapter 7), and sequenced action (Chapter 8). Chapters roughly follow the sequence a city will build incrementally. It is impossible to price what is not measured, enforce what is not observed, or design for outcomes that are not clearly articulated.