The OECD’s International Transport Forum has released a series of studies with exciting new ideas about how to reshape urban mobility, incorporating modern technology and new business and regulatory frameworks. The Fuse spoke with Philippe Crist, Project Manager for OECD’s research, about their latest findings.
Hayward: We’re seeing some sweeping trends in urban mobility—increased urbanization leading to surging congestion and demand, the rise of shared mobility systems, and in some cities, rather poor infrastructure. Tell us about the work you’ve done around new transportation systems that would solve some of these issues?
Crist: We have been looking at different scenarios and doing some outside-the-box thinking, and in our latest research, we’ve come up with an all-in shared fleet of vehicles that—in addition to a metro system—could comprehensively meet a city’s demand. What if you had every trip in a city replaced by a fleet of optimized, shared vehicles that could be either driving or self-driving vehicles? In our modeling, you would reduce the vehicle fleet by 97 percent, and cut vehicle kilometers travelled and energy consumption by 30-40 percent, and it’s possible with or without automation of the fleet.
In our modeling, you would reduce the vehicle fleet by 97 percent, and cut vehicle kilometers travelled and energy consumption by 30-40 percent, and it’s possible with or without automation of the fleet.
We took a mid-sized European city with good data on transportation patterns, and tried to imagine a system which could pick up or drop off people within a reasonable time frame, and offered service door to door. What we have found is that theoretically, you could replace every car and truck with about 3 percent of the vehicles as you have now, and offer essentially the same system performance. This system combines two levels of service; instantaneous app-based booking of shared “taxis” and a “taxi-bus” service, bookable 30 minutes in advance and servicing non-transfer trips from pop-up stops. Citizens communicate with a centralized dispatch and either get picked up where they are for the shared taxis or are directed to a pop-up station no more than 300 meters away for the “taxi-bus” service. Both services pick up and drop off passengers along the ride but respect the timing of the “original” trips they are replacing.
In our modeling, this system enables a huge reduction in the fleet of vehicles for the same quality of service—people in busses get better service, people in cars aren’t wasting fuel and road space by travelling alone anymore.
With the kind of system we modeled, you see that there’s a uniform coverage of high quality transportation access across the entire city, giving access that is just like everyone having a personal car, but without all the drawbacks.
In this particular kind of all-in scenario the potential for fleet reduction impacts are massive. In addition to taking so many vehicles off the road, there’s a 30-40 percent reduction in vehicle kilometers travelled, and a 30-40 percent reduction in CO2—without even electrifying the fleet. All of these gains occur simply based on the optimization of the routing, and the sharing aspect. Use of autonomous vehicles would decrease costs, but it doesn’t have a huge impact on the fleet or the vehicle kilometers travelled or the CO2 emissions. This is important, because it means we don’t have to wait for autonomous vehicle technology to be ready before we start optimizing urban traffic in this way.
This system also dramatically increases system performance and access. One of the things we saw was that for each 200×200 meter block of the city, under the current system, poorer areas of the city can only access 25 percent of the jobs in the city within 30 minutes of travel. The system is good along the major corridors and then it drops off precipitously. With the kind of system we modeled, you see that there’s a uniform coverage of high quality transportation access across the entire city, giving access that is just like everyone having a personal car, but without all the drawbacks. Door to door, this system provides better access at a better price than current public transport.
This system would also be extremely cost effective. In our calculations, the service doesn’t require any kind of subsidy, and the cost to the municipality is about 3/4ths less than current transit systems. It would be about a quarter of the cost to take a taxi, and half the cost to take a bus or public transport. So the outcomes are quite positive—but we can never have a system like this if we keep the same rules applying to transit and to taxis because this service will compete with both, and you can expect to get a lot of pushback.
Tell us more about the regulatory challenges.
If you want to make sure that this technology is applied in ways that decrease environmental and energy impacts and improve consumer outcomes, you have to think about how you define public transit. There are no busses in this scenario that we modeled, at least not traditional ones. There are no routes, there are no standard stops: Everything is optimized in real times and the stops are pop-up stops. We are agnostic about if the service is provided by a public or private operator, but the notion of public transport will have to change completely. There’s no theoretical reason why WMATA wouldn’t be able to provide these services… Or BMW. But under current regulations, neither one could provide these services—it doesn’t fit existing models, or the concession rules that you have to operate under, or the type of guarantees you have to provide back to the federal government, so we have to rethink the legal framework.
Are there other challenges to the transition, besides regulatory hurdles?
The situation I just described is one where everything shifts immediately at one point in time—we haven’t modeled how we get there, and the transition is a major vulnerability. We did some early testing of transitions and we find that it’s actually quite difficult to get the type of benefits we were modeling with 2-3 operators in same market, each optimizing a separate fleet for various routes and competing for the same trips. This challenge is avoided with a single dispatcher. But unless cities are really thinking ahead, they might not want to have just one operator due to concerns about competition—they might want several operators present in the market to make sure consumers are protected. In that scenario, you may have some consumer benefits, but not necessarily the urban or city benefits we have modeled thus far.
Second, even if we are only using a single operator to manage the new system, when we look at the transition towards our shared system with a lot of legacy cars on the road, it’s not clear we get the same kinds of efficiency gains. Travel distances for the shared cars become much further, and it becomes a very different optimization process.
Based on our modeling, once you have a 60 percent penetration rate where people start to switch to the centralized and shared system, we start to see these benefits.
Based on our modeling, once you have a 60 percent penetration rate where people start to switch to the centralized and shared system, we start to see these benefits. Up until then, it’s less efficient—you get more travel. That’s a challenge for a city looking to make the jump.
What about cultural challenges?
The biggest wild card, the place where it can all go wrong, is if people aren’t willing to share rides. In this system, all trips in the city are shared. It would be more convenient than the bus system we have now, which is very inefficient during off-peak hours and in the peripheral zones of cities. But for people who currently drive their own cars this involves a huge change in behavior, and we can’t anticipate its plausibility at this time. It’s possible that younger generations are much more comfortable with shared mobility and, given amenities like free Wi-Fi, they might even prefer it. People have disliked sharing in the past so future interest in sharing is an unknown variable.
Another issue is data privacy. The system we designed is very sensitive to tweaks in the algorithm and underlying assumption and it’s heavily reliant on good data to achieve efficiency gains. There has to be a new way in which to access information without having it available for nefarious purposes, requiring a new approach in how we interact with people and data in the private sector. The knowledge about how people are moving about on city streets is no longer held by public services, it’s held by private operators. We have to think of a way that the city can still function and get information from that data without compromising those companies.
Do you see any ways to shepherd electrification into this kind of system?
Yes. We looked at the impact on the fleet requirements if the city ran on electric vehicles, which means you have to account for shorter vehicle range, as well as longer recharging time. What we saw is that with a fairly realistic range (approximately 200 miles), and with a fairly short charging cycle (30-45 minutes, which is viable in the near future), you only need 2 percent more vehicles in your fleet to deliver the same service. It’s not an impediment.
With electric vehicles, what we saw is that with a fairly realistic range (approximately 200 miles), and with a fairly short charging cycle (30-45 minutes, which is viable in the near future), you only need 2 percent more vehicles in your fleet to deliver the same service. It’s not an impediment.
Perhaps more importantly, these vehicles will be used a lot more than vehicles currently owned by the public. Compared to current usage around 15-30 minutes a day, vehicles in a single shared system would be used 10-11 hours a day, which means that vehicles would wear out much more quickly. This could enable more rapid penetration of cleaner vehicle technologies—like electric vehicles. There’s also the benefits for fleet managers that come with the distribution of higher up-front costs for electric vehicles and their batteries amongst much greater users and distances travelled. If you are a motorist using an EV like a normal car it’s very costly per minute of use, but here your per-mile cost falls dramatically, and it’s much more compelling.
We’re also doing work on the specific aspect of shared vehicles, but we can’t look at these technologies coming in without looking at other trends and other technologies, so we are looking at regulations of commercial transportation service apps. One thing we know that will happen is that the regulatory framework for public transit and for taxis will have to change.
So what you’re saying is that if a city wanted to achieve these benefits, it would need to undertake a hard reset of the regulatory framework for taxis and public transit?
Yes, you need a hard reset. In our view, this is an issue for both commercial transit apps (like Uber or Lyft) as well as for a reworked urban mobility system. When we look at what we want to get out of the system and the regulatory tools that are currently in place, these two things are very poorly aligned. Right now, we have quotas, number of taxis, and other regulations—all very blunt instruments to get what we want: Competition, adequate coverage in terms of both hours of day and regions of the city, safety, etc. By using these blunt instruments to refine the system, we ultimately built in in a lot of inefficiency and failed to supply the true demand for taxi-like services, which is why we have seen such sweeping impacts from Uber and Lyft—the existing system wasn’t reaching its potential and was losing all those trips.
For all of these concerns, new app-based platforms can provide highly detailed data that allows authorities to monitor system performance. If we can get data on all those outcomes, we don’t need burdensome rules, and can put in place lighter regulatory frameworks.
Now, we have the capacity for much finer disaggregate data about vehicle routes and operation. We can imagine governments lightening the regulatory burden for both new entrants and incumbents in return for data that ensures authorities that public policy objectives are being met. That’s how you can get industry’s buy-in. We tell them, we have regulated for-hire passenger services for decades with little and infrequent feedback on the impacts of these services in our cities and on our streets. What should matter to public authorities—and to citizens—are not the rules themselves but the outcomes they ensure. Here, we are just trying to make sure that we don’t have bad outcomes—like intense competition in certain parts of the city, lack of access in other parts of the city, unsafe drivers or passengers being gouged—or unacceptable levels of congestion. For all of these concerns, new app-based platforms can provide highly detailed data that allows authorities to monitor system performance. If we can get data on all those outcomes, we don’t need burdensome rules, and can put in place lighter regulatory frameworks. This may also open the door for some of these companies to operate as quasi-public transit operators. Why could Lyft Line and Uber Pool not be considered part of a public transit network in a city with low density areas and off-peak hours? These services probably make more sense than running empty busses. So the hard reset is not just on the rules for public transit and taxis, it’s also on how we think about how the city controls movement and mobility overall.
What are other policy levers can cities use to enable a positive transition to a futuristic mobility system, rather than an unhelpful one?
Lack of coherent policy can create a huge mess. With autonomous vehicles, there’s potential that we will see these technologies deployed and they will have a huge impact on our cities and urban areas in general, but we don’t know yet what that will look like. But we can imagine that if policy isn’t applied correctly, we might see benefits in safety and convenience, but we could also see a dramatic increase in sprawl, or other inefficient uses of road space and land.
Increased inefficiency is a bad and costly outcome. When we think about that scenario, we can imagine people having multiple cars and sending empty cars on multiple errands simultaneously, or letting their car circle the block endlessly because the price of fuel is less than the price of parking. One way that cities could potentially prevent this is by pricing empty kilometers. If you have a system where you allow use of roads by empty vehicles to be less expensive than parking those vehicles in the cities, people will use the road to store their vehicle.
To prevent that, you need data, and you need to understand how these vehicles are being used and what’s happening to them. So that is another issue—because companies and citizens are not legally required to share data. It’s one of many examples where we need industry and government to work together, because this technology has the potential to change everything, in ways we cannot yet imagine.