Let’s take advantage of an article published in Le Monde on 9 March 2021 which reports the results of a study financed by the Forum Vies Mobiles, the SNCF think-tank, to look at the environmental footprint of autonomous mobility.
It is important to understand the functionalities and technical challenges associated with each level of autonomy and to distinguish several timeframes in the deployment of these systems. The Ecological Factory report predicts a massive arrival of autonomous mobility after 2050, i.e. too late to contribute to the SNBC’s objectives of decarbonising mobility. This reflects an amalgam between the different levels of autonomy because, while level 5 autonomous mobility systems will only be deployed in the long term, level 3 and 4 autonomous mobility services will be marketed before 2025.
Regarding connectivity, it should also be noted that while 5G will enable new applications, it is not essential for the deployment of many Level 3 and 4 autonomous services, as evidenced by all the Level 4 robot taxi services currently being deployed by Waymo in the Phoenix suburbs, the Level 3 personal vehicles planned by Honda and Mercedes in 2021, as well as the many autonomous shuttle experiments currently underway, which do not use 5G.
Regarding the environmental footprint of equipment, it is necessary to take into account all the components of mobility systems (vehicles and infrastructure) and to consider the systems deployed on the infrastructure as shared by all road users. This pooling of equipment could eventually lead to an overall rationalisation of the number and environmental footprint of the equipment required for an autonomous mobility service compared to the “all on-board” approach historically adopted by the automotive industry.
Concerning the energy costs of managing the data generated by autonomous vehicles, it is important to distinguish between the raw data processed locally in the vehicle and the less voluminous interpreted data, some of which could be sent to central servers by manufacturers or mobility operators.
It should also be appreciated that the massive need for driving data required during the development and validation phase of the systems (cf. the tens of millions of kilometres driven by Waymo on open roads) is out of all proportion to the volume of data generated by each vehicle in the long term and useful for their individual operation.
Contrary to the prediction of the Ecological Factory, the UN ECE Automated Lane Keeping Systems (ALKS) regulation of June 2020 will not limit future deployments by imposing adaptations to the road network; on the contrary, it will free up the marketing of certain systems that have been ready in the manufacturers’ toolboxes for years (Traffic Jam Pilot functions on motorways), without modifying the infrastructure. New regulations will then be introduced to govern the marketing of future level 3 and level 4 functions.
Deployment environments and scenarios
Concerning the reduction of emissions in urban areas, studies by the OECD’s International Transport Forum (ITF) have shown since 2015 that all journeys made by private vehicles in several major European cities could be made by shared robotaxis, representing only 10% of the current vehicle fleet, when supplemented by mass transit. This elimination of 90% of vehicles on the road in cities, although theoretical, confirms the strong potential of autonomous mobility to contribute to the decarbonisation of urban mobility.
Concerning the benefits of autonomous mobility services in rural areas, the first two autonomous mobility services in rural areas deployed in France give us a glimpse of what the use of technology in sparsely populated areas will bring to populations far from mobility: the Tornado project led by Renault in the vicinity of Rambouillet since 2017, and the autonomous shuttle service in the Val de Drôme led by Bertolami with Navya and Eurovia since September 2020.
Regarding deployment scenarios, the Ecological Factory report contrasts three scenarios: individual autonomous vehicles driven by car manufacturers, robot taxis driven by digital players, and autonomous public transport services driven by mobility operators. This simplistic vision of future mobility uses tends to ignore the fact that the mobility offer is already multiple. Individual vehicles, VTC or on-demand carpooling platforms, and public transport all have their place today in the mobility offer. In a context of continuously increasing demand for mobility, it seems difficult to imagine that automated versions of these different modes, which alone hold out many promises for decarbonising mobility, would not have their place in the panorama of tomorrow’s mobility offers.
A rigorous analysis of the technical, regulatory and service aspects is necessary to assess the benefits of the various forms of autonomous mobility. The net impact on the reduction of CO2 emissions could be positive if the increase in the footprint of in-vehicle systems is offset by the operational gains and virtuous uses made possible by autonomous mobility.
Autonomous mobility must therefore be conceived without any anti-technology demagogy, as a tool for improving our current modes of mobility, to make them cleaner, safer, more efficient and economically viable. The objective of decarbonising mobility by 2050 can only be achieved by combining new mobility services (carpooling, carsharing, bus services on motorways, etc.) with strong technological levers (decarbonised engines, alternative fuels, connectivity, automation, etc.).
Pierre Delaigue, Director of Connected, Autonomous and Electric Mobility projects, Leonard