What would disprove this analysis? (Criterion 1)

Mixed human-AV traffic operates safely without dedicated infrastructure investment

What would disprove this analysis? (Criterion 2)

AV adoption follows a rapid S-curve that minimizes the transition period

What would disprove this analysis? (Criterion 3)

Zero-occupancy vehicle miles remain below 10% of total AV miles traveled

When should you avoid autonomous vehicle infrastructure redesign?

You're making 50-year infrastructure investments based on uncertain AV adoption timelines. You assume AVs will eliminate the need for parking within the next decade

What are alternatives?

Adaptive infrastructure: Modular designs that evolve with adoption rates. Transit-first investment: Public transit improvements work regardless of AV timeline. Congestion pricing: Manage demand rather than building for uncertain supply changes

Catalog

I009

Infrastructure

Autonomous Vehicle Infrastructure Redesign

MEDIUM(75%)

February 2026

3 sources

Context

Autonomous vehicles promise to reduce accidents, eliminate parking needs, and optimize traffic flow. Cities plan for this future by investing in smart infrastructure, V2X communication systems, and AV-ready road designs. But the transition period — decades of mixed human and autonomous traffic — creates infrastructure demands that neither system was designed for. Roads need both traditional signage and digital communication. Parking structures built for human drivers become stranded assets while new pickup/dropoff zones create congestion. The infrastructure investment required for the transition may exceed the cost of the final autonomous state, and cities that invest early risk building for a future that arrives differently than planned.

Hypothesis

What people believe

“Self-driving cars will reduce accidents and transform urban infrastructure for the better.”

Actual Chain

→

Mixed traffic period lasts decades(Human and AV traffic coexist poorly)

└

Infrastructure must serve both paradigms simultaneously

└

AV-optimized roads confuse human drivers

└

Transition costs exceed steady-state savings

→

Parking infrastructure becomes stranded assets($200B+ in US parking structures)

└

Garages built for 50-year lifespan become obsolete in 20

└

Urban land locked in parking cannot be repurposed quickly

→

Pickup/dropoff zones create new congestion(Curb space becomes scarce resource)

└

Double-parking by AVs waiting for passengers

└

Zero-occupancy vehicle miles increase as AVs circle

Impact

Metric	Before	After	Delta
Transition infrastructure cost	Expected: offset by savings	2-3x steady-state cost	+100-200%
Parking asset stranding risk	$200B+ invested	Potentially obsolete	Massive write-down
Zero-occupancy vehicle miles	0%	Up to 40% of AV miles	+40%

Navigation

Don't If

•You're making 50-year infrastructure investments based on uncertain AV adoption timelines
•You assume AVs will eliminate the need for parking within the next decade

If You Must

1.Design flexible infrastructure that serves both human and autonomous traffic
2.Build parking structures that can be converted to other uses
3.Invest in curb management systems before AV adoption reaches critical mass

Alternatives

Adaptive infrastructure — Modular designs that evolve with adoption rates
Transit-first investment — Public transit improvements work regardless of AV timeline
Congestion pricing — Manage demand rather than building for uncertain supply changes

Falsifiability

This analysis is wrong if:

Mixed human-AV traffic operates safely without dedicated infrastructure investment
AV adoption follows a rapid S-curve that minimizes the transition period
Zero-occupancy vehicle miles remain below 10% of total AV miles traveled

Sources

1.
RAND Corporation: Autonomous Vehicle Policy
Analysis of transition period challenges and infrastructure costs
2.
NACTO: Blueprint for Autonomous Urbanism
Urban planning framework for AV transition
3.
UC Davis: Three Revolutions in Urban Transportation
Research on zero-occupancy miles and induced demand from AVs

I003 I010