Nouvelles mobilités

NFM development : local challenges, global actors – an illustrative case, the robotaxis

One of the studies led by YKems in 2018 aimed at qualifying the impact of a completely new range of NFM on Lyon metropolis (France) modal shares. The results obtained showed that, in the long run, the robotaxis could claim, depending on the geographical area covered and the level of car sharing, between 19 and 25% of daily trips, at the expense of private cars (PC), public transport (PT) and other NFMs.

From the industrial economy standpoint and its approach of the evolution of value-chains while confronted to radical innovations, several phenomena are interesting in robotaxis at both global and local levels:

• A small number of international players (Waymo, Zoox, GM / Cruise, Uber / Toyota …), with access to financing and fitted with deep pockets, are aiming straight at a global market for robotaxis. They seem to be in a “winner takes all” or “takes most” logic, willing to amortize the considerable expenses in the R&D race and looking for simultaneously capturing the customer through the multimodal interface (which gives sense to Mobility as-a-Service) and building an unquestionable market power towards subcontractors (car and equipment manufacturers).

• A development in direct competition with the historical mobility, depending on national components (highway code, safety standards, labor costs, etc.) but also resolutely local ones, each and every city in the world having its particular characteristics: more or less important density, more or less developed and efficient public transport, more or less saturated road networks etc. The transport policy is in the hands of local authorities who intervene via public service delegations (PSD) to address the mobility needs, but also to handle issues such as pollution, congestion, traffic conditions, parking offer etc. all these questions being part of a broader one which is how to use common goods and reduce negative externalities.

Will the firms aspiring to a dominant market share worldwide have the capacity to negotiate with each city, when local projects depend on the political agenda and the citizens expectations? Should they, for example, partner with companies that have historically developed their expertise in this field (Keolis, Transdev, RATPdev and other local public transport operators)? What place on the mobility value chain will remain to be taken by companies investing and employing at national and local levels in this announced mobility revolution?

Local issues of robotaxis

If the “endgame” impact of robotaxis seems to be clear, the trajectory to achieve it is much less obvious. Among the barriers to implement such fleets, the local Public Transport Authorities (PTAs) would have to be convinced of the usefulness of robotaxis as a complement to public transport and not as a supplement to the private vehicles already on the road. Yet, the fear of cities to see congestion increasing because of robotaxis is real, and arguing in favor of robotaxis sharing is not always enough to convince them. In addition, the “live” experiments carried out in France are either on too small a scale – and therefore not much convincing – or unfeasable in the existing flow for obvious problems of regulation and security.

Before launching costly experimentations, and in order to choose the best zones and routes for robotaxis’ development, performing large scale simulations at city level would be a prerequisite to demonstrate the utility, the feasibility and the optimal sizing of an efficient robotaxis fleet. Such analyses could be made using models allowing to combine three levels of analysis:

• A modelling of a given city and its travel needs (based on ABM type tools);
• An in-depth and quantified analysis of the competitive dynamics between the different mobility modes allowing to assess potential market shares of alternative solutions;

An economic analysis to assess the interest of PTAs in fostering transport modes, according to the infrastructure investments they would require (reserved lanes, connected roads, etc …)

Local adoption of robotaxis will also depend on PTA governance. Current PSDs, mostly exclusively dedicated to public transports, will have to expand their scope to robotaxis, especially if PTAs wish to develop the intermodality between these two transport modes. The alternative would be to let robotaxi companies operate freely, like their free-floating bicycles counterparts, and regulate later. But such an approach seems far too risky for the common good. The issue here would then be to know how the PTA will adapt the PSD/Operators system to integrate the governance of new transport modes such as robotaxis, while keeping a coherent public transport policy.

Global issues of robotaxis

This study of new autonomous vehicle segments includes an analysis of the associated costs and revenues. For example, droids’ and lockers’ logistics are deemed cheaper to produce and to operate than the robotaxis, while recent studies assert the data storage and processing costs for robotaxis are higher than previously reported. Finally, the revenues associated with data generated by autonomous vehicles and their passengers also deserve a re-estimation, beyond the mere price of the ride. Indeed, without understanding the income generated by robotaxis operators, it is impossible for PTAs to tax them at a fair price for the usage they will make of the common good, i.e. the road. In this area, the work of regulatory economists (cf Jean Tirole²) on this topic will allow to frame the reflection and to release specific principles for robotaxis (contracts limiting the opportunistic behaviors of the regulator and the regulated, pricing of the access to the road).

Beyond the local issue of robotaxis implementation, if new actors like Waymo massively invest in this technology, it is also the case of car and equipment manufacturers (in different forms of alliance with application editors, suppliers specialized in sensors and processors, etc.) and transport operators³, with repercussions on their respective positioning on the mobility value chain.

[1] Agent-Based Modeling

[2] See his educational booklet « Economy of the Common Good », pages 604-628, PUF 2016.

[3] Partenership Transdev/ZF ragarding autonomous shuttles, as well as Valeo and Keolis with Navya.