Self-driving car and the future of communities

Self-driving car, city planning and the future of communities seen with… “Steer the autonomous vehicle to serve the city” from Yann Leriche and Jean-Pierre Orfeuil

?

I was lucky enough to have a manager who would regularly offer us some essays on mobility. In 2019 he provided me with the book from Yann Leriche (at the time, Transdev North America CEO and head of autonomous driving program at Transdev) and Jean-Pierre Orfeuil (renowned urban planning expert) “Steer the autonomous vehicle to serve the city”.

?

Preparing the way

The authors recall that self-driving car has been long to take a place in the public debate and imagination, as OEM had chosen to pursue their effort on driving assistance and emphasize the pleasure of driving. The development of Google’s Waymo or Tesla technologies, thanks to tremendous investment as well as a strong vision and storytelling has changed the game.

The book is organized around 10 chapters which define two main parts.

In the first one, which goes from chapter 1 to 4, the authors give an overview of self-driving cars innovations, how it managed to catch the hype, the hopes, and expectations raised, as well as the possible risks one might identified, and the critics formulated by some.

Second part focuses on the proposition the 2 authors push forward: the “TRUST” framework. TRUST is their propositions to set boundaries around the development of self-driving car, so that this disruption in mobility is put at the service of communities, would it be in cities or in more rural areas, in the frame of global climate change challenges.

This article will focus on the TRUST framework as formulated in the book. ??

The “TRUST” framework stands for:

• Technology
• Rules
• Usages
• Services
• Territories (communities)

Authors underline that the conjunction of these 5 topics forming the acronym trust was done on purpose. Trust is what is and has always been necessary to build society and is at the base of our capability to leave together as humans being as soon as we step out of the family clan.

What are the stakes and issues the two authors identify for each topic?

?

- Technologies -

The authors address this subject through 3 points of view.

First, they set the discussion around the technologies used to make the autonomous driving possible: sensors and algorithm that allow the robot-car to sense – plan – act. They give a quick overview of the different technologies used: radar, lidar, sonar or GPS to sense ; ruled-based or machine-learning algorithm and computing to plan ; command technologies (brakes, gear, steer…) to act. They emphasize both the ancientness of these technologies and the recent and rapid progress they have seen. They make a special focus over algorithm, underlying the challenges to define computing method able to tackle the inherent complexity and uncertainness of any driving environment.

The second point of view tackles the price issues. These technologies have a cost, and it is all the higher as you are aiming for a higher degree of autonomy. The authors mention some research paper mentioning $100k additional cost compared to a classical car; as well as some industrial positions referring to limited extra-cost in the near future, with the technology and the industrial process reaching maturity. The main issue is to build a vision over little knowledge of the final economic impact over the cost of the vehicle (acquisition and maintenance), but also cost of the infrastructure, impact on insurance policies, accidentology cost…etc.

Third point to address is the expected benefit of these technologies. They are multiple: end of driving license barrier, reallocation of driving time to higher-value activities, data valorization, road network optimization and increased capacity through cruise control and improved road safety.

This first chapter end on some thoughts over the challenges and main problem to address, with 3 main concerns: 1. If the final goal is well defined and appealing, the path to it is not so clear and raise many questions regarding the cost and immediate benefits ; 2. How will we ever be sure that the artificial intelligence driving the car is up to our expectations regarding safety, including wider issues as cyber-war or cyber-terrorism ; 3. Are our current political and administrative process ready to tackle the introduction of robot-car in our streets and communities ?

This last question is in direct connection with the next chapter: Rules.

- Rules -

This chapter starts with some memories of how the increase of hypomobile traffic in Paris in the early years of the 19th century led to the instauration of new rules that basically set a large part of how cities as me know them today in France and Europe are structure: avoidance of road occupancy for private motives, streets and facades in straight lines, invention of the side-walk and obligation to use them for walking… later, with the arrival of the motor-car it will be traffic signs and lights, speed limits and so on.

The 2 authors also remind us that like previous new technologies, driverless-vehicles are in this “early-morning moment” where their downsides are completely hidden by the promises of better future they enclose.

And yet, the roll-out of a driverless mobility system cannot be directly compared as the deployment of smartphone and internet usage, because most of its impact will occur not in the cyberspace but in the real world where constant frictions, scarcity of space, usage rivalry and competition, positive marginal-cost and physical dangers, including lethal ones, occur.

The authors identified several regulation issues that will need to be addressed with the driverless vehicles:

• Homologation: the main issue identified is the establishment of a trusted tier responsible for homologation and able to deal with non rule-based algorithm, in constant evolution. ??
• Road code, street usage: the authors give some example of immediate thought issues: how to deal with a pedestrian who could freeze an AV on a pedestrian crossing just by standing-by in front of it? How do we regulate a potentially anarchic AV last mile delivery system?
• Ethic: the “tramway dilemma” of Philippa Ruth Foot will apply, and all the more as the ultimate decision will be made by a robot. What ethical basic regulation will apply to all interfacial intelligence and what will be left to the decision of manufacturer or local communities?
• Data: autonomous vehicles will collect multiple data: owner and user data, vehicle data, localization data and environment data. This vast amount of data pose question over privacy protection but also their usage and to the benefit of whom. The authors remind us so with the issues of personal GPS optimizing your personal journey to the detriment of the collective interest (secondary roads alternative route).
• Finance: current economical and finance flow should largely be disrupted by the introduction of electric driverless car: change in ownership habits, less accident, less penalties, less insurance policies… on the other hands, il will also foster some new investments and needs. ?
• Planning: local public authorities will be at the forefront to set the rules and find the right urban planification, support the evolution of land use (parking, charger, flow…) and set the cohabitation between pedestrian, bike-riders and self-driving cars, not to mention the intermediate period where AV will cohabit with usual car.

- Usages -

What is at stakes according to the two authors is a possible revolution in the use of our time-budget allowance for transport.

The authors take the assumptions that we have reached a moment where autonomous driving is technically possible and at an acceptable cost. In this hypothesis, it will replace or complement the already available transport solutions for individuals, companies and public authorities.

Choice we make over public transport means are driven by different factors, in order: cost, time, experience, conviction.

• Cost: with no driver needed, the difference between a taxi and your personal car is erased. Therefore, the main driver to the development of robot-taxi services will be the attitude towards shared assets and the cost of operation management vs. personal owning.
• Time: Not only driverless vehicle could improve the time needed for one given trip, by optimizing road debit; it will also allow to avoid some of the side-activities to driving (fuel, parking…).
• Experience: how much the fact we do not have to drive anymore will modify our time allocation of transport (Zarawi constant of 1h) as we are allowed to save time on transport (cf. supra) but also to focus on other, more added-value, task than driving?
• Convictions: Choice of transport mode relies on objective, quantitative criteria (time, price…) but also on social belief, habits and norms. How much are we ready to share one’s vehicle and or one’s journey with a complete stranger? ?

?On all of the first 3 aspects, all projections also depend on induced bounce-back effect, that might annihilate the initials benefits brought.

?As a temporary conclusion, the authors see two options regarding usage.

In option 1, vehicle-as-a-service has overcome all other form of usage and ownership. It is a bold option as the development of the economy of sharing is yet to be confirmed. In option 2, vehicle remain privately owned. AV owner leverage the optimization of time allowed by the autonomous driving, which then compensate the overset of owning the vehicle; sharing practices possibly emerge in the network of close relatives. ???

?

- Services and Systems -

In direct connection with the previous chapter, this part is about how the autonomous driving will shape and change the current mobility services and systems.

Car exceeds all other modes as the main transport mode, the benefit it offers as a transport mode but also as an economical asset and as an extension of your privacy are unrivalled. Roughly, 80% of all distance are made by car in France, for example. Noinetheless, car is in in the hot seat, especially in most dense, metropolitan, area for its downsides, especially road congestion and land use. It is not very clear that AV could solve these issues. According to the authors, it will greatly depend on the usages (private or as-a-service) and the incentives set on each territory to massify transport flows.

Public transit on the other hand is suffering the Baumol’s cost disease like other many services public sector. There is little possibility of productivity gain in the way a network is operated and payroll account for 75% of the cost of service. Driverless vehicles offer both the possibility to reduce the cost of transit network and to dramatically transform the way mobility services are provided on all part of the area, considering form example the development of robot taxi-fleet, as part of mobility-as-a-Service offer.

Whatever the options that will overcome, current systems and services can be completely turned around by autonomous vehicle: from vehicle design and personalization to road segmentation and services, via portfolio of mobility services and hierarchy of modes, from first & last mile solutions to long distance travel and daily commute.

?

- Territories (communities) -

Driverless vehicle, if they hold their promises, will reduce the cost of transport and thus, increase the options for mobility. The impact is increased land sprawling and distances between activities.

How will the different level of our communities be affected by the development and roll-out of AV?

The authors start to review some past innovations and the impact they effectively had on the way we live (and the impact they eventually did not have…):

• Elevator, which did change the way we build business district but did not really transform the city into vertical spaces as individual inspirations go for the individual family house
• Public transit, which initially developed in central business area of European cities as a private, competitive service and evolve under regulation to a mean of land planning and a way to connect workforce with factories.
• Sidewalk, which initially free the road for horse and then motor coaches before becoming a way to regain control over the speed and space usage of these same vehicles

Regarding AV, the authors differentiate local mobility, agglomeration mobility, metropolitan mobility in connection with long-distance travel.

Local mobility: the aim will be to keep with the reduction of spaces allocated to vehicles in favor of walking, biking and other individual micro mobility modes. In this regard, the use of AV will be dedicated to specific people (elderly) or to low-density area.

Agglomeration-level mobility: in a positive vision of the future, AV would both reduce the cost of transport (especially time budget, as one could do other things while being transport) and thus open new economic opportunities (work or study further away from home). The induced traffic would be counterbalanced by the optimization made possible by vehicle-to-vehicle and V2X connection. Nonetheless, the authors also acknowledge the need for transport, urban planning, and fiscal policy that limit the sprawling effect of such new possibility.

Metropolitan and intercity mobility: AV could contribute to the reinforcement (if they do not become a must…) of world-class metropolitan area, integrated in the international chains of exchanges. As they would enlarge the area of connection, for example through dedicated lane on highway allowing higher speed journey for people or goods, AV would contribute to concentrate different area of specialization and reinforce collaboration.

?

?Conclusion?

I leave the words of conclusion to the authors, in their call for citizens to size the stakes and challenges posed by the evolution of automotive world, in particular, and technology in a broader perspective.

“In this book we have wanted to drive attention on the risk that a fascination for technology might relegate what will always be the most important: the nature of the spaces we habit, the relationships we forged there, and the connections of all kinds they allow.

As soon as the self-driving car will leave their laboratories, test tracks or experimentation sites, it will become a matter of social interest, as the car has been before. One can hope, and even wish, that prepared public local authorities and well-informed citizens will know how to civilize it, sooner and better that they did for its ancestor.” ????

Once, I had attended a management seminar in Transdev. The consultants hired to animate this session said something about the change curve that stroke me: first come the technology, second social adoption and in the end, the regulation chase after the 2 other.

Let us hope that for autonomous vehicles, we manage to rule against the saying.?