Check out the episode:
You can find the shownotes through this link.
Are you interested in the effects of autonomous vehicles on the urban fabric?
Our debate today works with the article titled How autonomous vehicles can affect anomalies of urban transportation from 2025, by Francesco Filippi and Adriano Alessandrini, published in the MDPI Future Transportation journal.
This is a great preparation to our next interview with John Rossant in episode 440 talking about the mobility revolution involving autonomous vehicles.
Since we are investigating the future of cities, I thought it would be interesting to see how autonomous vehicles – AVs can revolutionise urban transport by addressing systemic issues like congestion and safety. This article argues that achieving great urban futures depends on integrated urban planning and robust policy regulation working with technology.
[intro music]
Welcome to today’s What is The Future For Cities podcast and its Research episode; my name is Fanni, and today we will introduce a research by summarising it. The episode really is just a short summary of the original investigation, and, in case it is interesting enough, I would encourage everyone to check out the whole documentation. This conversation was produced and generated with Notebook LM as two hosts dissecting the whole research.
[music]
Speaker 1: Imagine your city 10 years from now. Do self-driving cars just glide perfectly in unison down the highway, basically banishing traffic jams to the history books?
Speaker 2: Or have we accidentally designed this nightmare where empty ghost cars just circle the block 24 hours a day because driving around is actually cheaper than paying for a parking spot?
Speaker 1: It’s a pretty terrifying image. Today, we are looking at a really comprehensive academic assessment of how autonomous vehicles, AVs, are going to interact with traditional urban transport anomalies. The central question we are exploring is, will the technological advancement of AVs organically solve urban mobility crises like congestion and sprawl?
Speaker 2: Or will this technology actually make those exact problems worse unless we strictly confine them to a shared station-based paradigm?
Speaker 1: I will be arguing that the intrinsic technological efficiencies and, frankly, the safety breakthroughs of AVs will organically revolutionize urban transport across all use cases.
Speaker 2: And I will argue that without enforcing a massive paradigm shift towards shared autonomous electric vehicles, or SEVs, AVs are going to disastrously amplify urban sprawl, congestion, and public health issues.
Speaker 1: Let’s start with the technology itself because I think it’s crucial to understand the magnitude of the leap we are taking here. This isn’t just like advanced cruise control.
Speaker 2: No, definitely not.
Speaker 1: The data from real world deployments is historic. Take Waymo’s operations in California. They are handling over 700,000 monthly trips, over 56.7 million miles. They have documented a 96% decrease in intersection crashes compared to human benchmarks.
Speaker 2: That is an impressive statistic, absolutely.
Speaker 1: But it’s the way they drive too. An autonomous system has 360-degree radar. It calculates predictive braking in a fraction of a millisecond by smoothing out speed variations, eliminating all that erratic braking humans do They reduce fuel consumption by 15 to 20%. The technology inherently cures human physical inefficiency.
Speaker 2: I don’t disagree with the mechanics of the vehicle itself. The sensor arrays are brilliant. But you are looking at the efficiency of a single isolated trip and completely ignoring the macro system, specifically the trap of rebound effects.
Speaker 1: Rebound effects?
Speaker 2: Like induced demand. What happens to human behaviour when you make an activity completely effortless? The US EPA estimates that private AV use could actually increase overall energy consumption by up to 120%.
Speaker 1: Wait, let me make sure I’m hearing you. You’re saying a technology that drives 20% more efficiently is somehow going to more than double our energy use?
Speaker 2: Exactly, because of zero occupancy runs. Let’s say your AV drives you to work downtown. Currently, you’d pay $30 to park it there all day. But your AV doesn’t need to park. It just drops you off and drives itself all the way back to your house in the suburbs to park for free, and then it drives all the way back to pick you up at 5:00. You have just doubled the miles travelled for the exact same utility. The only viable solution in the literature is replacing privately owned AVs with a shared fleet mobility as a service. With dynamic routing, a shared fleet only requires 15 to 25% of the current number of vehicles to fulfill the same demand.
Speaker 1: I see why you think that, but isn’t this kind of upgrading a city from dial-up internet to fibre optic? Let me explain. I think you are underestimating the raw physics of AV platooning. Human traffic jams are a domino effect. One person taps their brakes, and a mile back, traffic stops.
Speaker 2: Phantom traffic jams, yes.
Speaker 1: But connected AVs don’t do that. When you have connected AVs, say four-meter vehicles with just a one-meter gap between them, they move as a single synchronized organism. Even at just 30 kilometres per hour, that setup can achieve a massive flow rate of 5,760 vehicles per hour per lane.
Speaker 2: That’s the theory anyway.
Speaker 1: It’s the physics. The pipeline becomes so fundamentally wider and faster that it naturally absorbs the extra demand without needing a centralized authority to ration the bandwidth.
Speaker 2: I’m sorry, but I just don’t buy that. That beautiful flock of birds metaphor only exists in a vacuum where 100% of the vehicles are perfectly connected. The literature shows that flow improvements evaporate without complete connectivity.
Speaker 1: Really? Evaporate completely?
Speaker 2: Yes. A traffic stream of 100% non-connected AVs, cars that drive themselves but don’t digitally talk to each other, can actually decrease lane capacity by 20%.
Speaker 1: Wait, because they are too cautious?
Speaker 2: Precisely. They operate in a sort of turtle mode. But honestly, even if we assume perfect platooning, you run straight into a wall called Marchetti’s constant.
Speaker 1: We’re talking about the human travel time budget?
Speaker 2: Exactly. Humanity has always maintained a daily travel time budget of about one hour, 30 minutes each way. If you look at an ancient Roman citizen walking to the forum, it took 30 minutes. An 1890s factory worker on a streetcar, 30 minutes. We don’t use faster transport to save time. We use it to travel further distances.
Speaker 1: So you’re saying if my AV platoon allows me to travel three times faster, I won’t just enjoy a 10-minute commute.
Speaker 2: Human history guarantees you won’t. You will move three times further away from the city centre to get a bigger house and commute for 30 minutes from there. Faster travel simply induces people to stretch the city out, sparking a 10 to 30% increase in urban sprawl.
Speaker 1: That’s an interesting point, though I would frame it differently. You’re assuming the distance someone travels is the only metric. Think about the profound social benefits of door-to-door AVs, especially for accessibility.
Speaker 2: Accessibility is important, yes.
Speaker 1: It’s vital. We have an epidemic of loneliness among elderly and disabled populations. Right now, paratransit takes days of advance booking. Giving them access to a door-to-door AV is like giving them a personal teleporter. It restores complete independence.
Speaker 2: I wanna be very clear here. For vulnerable groups, spontaneous point-to-point transit is an absolute necessity. AVs will be life-changing for them. But if we extend that door-to-door luxury to the general able-bodied public, we create a catastrophe for both urban congestion and public health.
Speaker 1: So what’s the alternative?
Speaker 2: For the general public, we must use the elevator model or a transit hub model. Station-to-station service requiring a five to 10-minute walk. If everyone has door-to-door AVs, imagine the curb side bottlenecks. You’d have 100 vehicles trying to pull over at the exact same coffee shop at 8:00 AM.
Speaker 1: I have to challenge that. By pushing for a station-based model that forces people to walk, aren’t we artificially hobbling the greatest mobility breakthrough of our century just to engineer incidental exercise?
Speaker 2: Because the public health data demands it. Look at the upcoming Trenton Moves project in New Jersey. They are deploying 100 autonomous electric shuttles and 50 kiosks. They strategically place them so 90% of residents are within a five-minute walk, balancing efficiency with health.
Speaker 1: I get the health argument, but AVs are gonna radically alter urban space for the better anyway. Because an AV can drop you off and park remotely, cities will inherently reclaim massive amounts of asphalt.
Speaker 2: If we manage it correctly.
Speaker 1: If an AV drops me off and immediately leaves to park 10 miles away, the city still gets to rip up the downtown parking lot and plant trees. Why does it matter if I own the car or rent it?
Speaker 2: It matters immensely because of the physical footprint of those empty miles. A privately owned vehicle parking 10 miles away means massive empty miles driven.
Speaker 1: But if my vehicle is electric, it doesn’t have a tailpipe, it isn’t burning gas. Why does it matter?
Speaker 2: Because of particulate matter pollution from tire and brake wear. Electric vehicles are heavy. When you combine that battery weight with instant torque, tires degrade faster. All those extra empty miles are physically shredding microscopic rubber into the air.
Speaker 1: But you just said earlier that the autonomous system operates smoothly. Doesn’t that minimize tire degradation?
Speaker 2: Minimized, maybe. But the sheer volume of extra miles completely overwhelms those gains. This is exactly why the ETH Zurich study is so critical. Their modelling states that individual traffic decline only occurs if private cars are limited to a maximum of 20% of the total fleet.
Speaker 1: The 20% solution.
Speaker 2: If we just let the consumer market run wild, reclaimed space won’t turn into parks, it will turn into wider roads for empty AVs. Reclaimed space must be deliberately engineered into diffused transit-oriented development, or TOD.
Speaker 1: Now, traditional TOD usually means high-density housing within walking distance of a major train station. How does diffused TOD change that?
Speaker 2: Diffused TOD uses a network of mobility hubs connected by shared AVs. The Vinnova project in Sweden is piloting exactly this. By using shared shuttles to solve the first-mile and last-mile problem, they extend the reach of a transit node. But it only works if the vehicles are shared.
Speaker 1: I see the logic in the Trenton and Vinnova models. But I keep returning to the staggering empirical data regarding safety. A recent study by Swiss Re and Waymo found an 88% reduction in property damage claims, and a 92% reduction in bodily injury claims compared to human drivers. That level of risk reduction is historic. My position remains that these safety and capacity improvements represent an organic immediate cure for urban transport woes. The market will optimize for efficiency, and the spatial benefits will naturally follow.
Speaker 2: And my position remains that hardware alone cannot beat human behaviour. Without strict mobility hubs, mass integration, and limiting private ownership, AVs will merely automate and accelerate resource depletion and urban sprawl.
Speaker 1: We do definitely find clear convergence on one major point today. Autonomous vehicles are going to fundamentally transform the physical landscape of our cities. We are going to reclaim vast amounts of urban space.
Speaker 2: Agreed. The parking lot as we know it is doomed. The lingering disagreement is just whether consumer-driven adoption is sufficient, or if strict service-based paradigm shifts are required to prevent a public infrastructure failure.
Speaker 1: It really highlights how deeply complex the intersection of technology, human behaviour, and urban planning truly is. We encourage our listeners to review the source material and see how these theories are playing out in real-world test cities today.
[music]
What is the future for cities podcast?
Episode and transcript generated with Descript assistance (affiliate link).
What is The Future for Cities? Podcast 
Leave a comment