Article
Autonomous Vehicles: How Can They Contribute to Decarbonizing Mobility?
Autonomous Vehicles: How Can They Contribute to Decarbonizing Mobility?
By Stéphane Amant, Senior Manager at Carbone 4
GLOSSARY
AOM = Mobility Organizing Authorities MA = Autonomous Mobility MAD = Sustainable Autonomous Mobility NTIC = New Information and Communication Technologies TC = Public Transportation VA = Autonomous Vehicle
Self-Driving Cars: Between Fantasy and Questions

Observers are convinced that autonomous vehicles will soon be a reality and will revolutionize the way we use cars. Increased safety, improved traffic flow, greater productivity for businesses, and more: the car of tomorrow seems to have it all and fuels all kinds of fantasies. However, this rosy picture is sorely lacking one major concern: Will the advent of autonomous vehicles contribute positively—or not—to the decarbonization of transportation, a key sustainability issue involving the reduction of oil consumption? The answer to this key question—given how necessary the transition to a low-carbon economy is—is by no means clear at this point. Indeed, the rise of autonomous vehicles (whatever form they may take: private cars, robot taxis, shuttles) raises a number of questions.
Rebound Effect and Increased Preference for Cars
Admittedly, the electrification of powertrains and the automation of driving (which could promote eco-driving and make “road trains” possible) are all aspects that, at first glance, appear to be positive for decarbonization. Conversely, if we consider the issue from a more systemic perspective, won’t the development of this type of vehicle significantly reinforce the public’s preference for cars? After all, for many households, having a vehicle that allows them to get around independently while doing something else at the same time could be very appealing. This could then result not only in a rebound effect in individual car use (more kilometers traveled due to increased travel or urban sprawl), with more vehicles on the roads, but also due to reduced use of competing modes such as active modes of transportation (walking, biking, scooters, etc.) or public transit (train, tram, bus, coach, etc.), which are, in principle, more energy-efficient. This represents a potential obstacle to the rapid decarbonization of transportation.
The Energy Impact of Digital Technology
Finally, from a purely technical standpoint, the autonomous mobility system relies on a plethora of electronic equipment. However, the role of information and communication technologies (ICT) in the energy transition remains highly ambiguous today. Research by The Shift Project [1] on digital minimalism provide evidence of this: The impacts of digital technology are still underestimated, with a clear increase in the environmental footprint in most cases. As a reminder, it is estimated that Digital activities accounted for 3% of GHG emissions in 2017, which is as much as air travel. Based on current trends, by as early as 2025, this figure could rise to 7 to 8% of the total—equal to the combined share of the air and maritime transportation sectors at that time! In fact, we almost always overlook the very material aspect of the digital ecosystem: the construction and operation of networks, data centers, and devices (such as our smartphones or computers), as well as sensors in the case of the Internet of Things (IoT). Behind a world perceived by authorities and individuals as intangible lies, in fact, a great deal of embodied energy from manufacturing and a great deal of operational energy. The indispensable role of digital technology in the implementation of autonomous vehicles could therefore have negative repercussions in terms of decarbonization. Although not exhaustive, these questions provide an initial glimpse into the research needed to better assess the future role of autonomous vehicles in the energy consumption required for our mobility.
Various Possible Models for Autonomous Mobility
So how can we know whether we’ll be able to combine MA with the decarbonization of mobility in the future? To try to answer this question, we must first and foremost determine which MA we’re talking about. However, the future of MA is still very uncertain. Several paths are possible depending on the organizational choices that are made, each with its own risks and benefits in terms of sustainability (see a recent IDDRI publication on the subject[2]). Three MA models illustrate these different possibilities :
- Model M1: dominated by the individual private autonomous mobility, where automakers play a central role
- Model M2: dominated by the autonomous, collective, and shared mobility, where local governments (AOM) and public transit operators play a central role
- Model M3: dominated by the autonomous mobility focused on private services of robot taxis, where digital companies (such as the GAFA) play a central role
Clearly, the benefits and risks in terms of decarbonization are not the same across these three MA models. For M1, the main advantage from the manufacturers’ perspective is that it would require fewer changes for drivers, thereby ensuring the sustainability of their business. On the other hand, we could anticipate a continuation of solo car use, a high risk of a rebound effect (linked to the possibility of using one’s time in multiple ways while in the vehicle), and significant upstream impacts related to the mass production of individual equipment. For Scenario M2, the appeal of public transit would be enhanced, which would, in principle, be more conducive to decarbonizing transportation use. On the other hand, we could anticipate difficulties in implementing shared mobility innovations by public sector actors who are not specialists in the field. For M3, by encouraging carsharing, the number of vehicles should be reduced, which would, in principle, be a step in the right direction for decarbonization. Nevertheless, questions regarding equitable access to services would arise immediately, as would likely issues of local governance with the AOMs. The future of MA will likely be a more or less balanced combination of these three models, of which M2 and M3 are certainly the most climate-friendly. However, the conditions for success still need to be clarified, which is why the AOMs must take responsibility for: (i) go beyond “mere” management of TCs to ensure coordination with private MA offerings; (ii) ensure consistency between the uses of these new MA technologies and the imperatives of sustainability, which are non-negotiable. At present, there is no magic formula: public actors can promote MAD either directly or indirectly.
- Directly, for example, by developing connected infrastructure and licensing its use to MA operators;
- Indirectly, by establishing rules that further legitimize the MAD (through parking, congestion pricing, Low-Emission Zones, etc.) and by launching public procurement contracts with stringent criteria for decarbonizing mobility.
On the Need to Adopt a Systems Approach to Impact Analysis
Our conclusion consists of three key messages. First, it is crucial not to focus on technologies, but rather to consider the uses first and foremost and to ensure consistency with urban and suburban planning. Next, it is essential to take advantage of MA's current experiments to develop and monitor sustainable mobility indicators. Finally, at a time when the fight against climate change is entering a critical phase, it is essential to conduct comprehensive analyses of the environmental impacts of these various experiments. Furthermore, we must now conduct a forward-looking analysis of the impacts of the various MA models proposed above (taking the underlying digital ecosystem explicitly into account). Only on this basis will we be able to identify the conditions necessary to ensure that the development of the MA is compatible with the pursuit of the public interest—namely, contributing to the rapid decarbonization of the economy.
[1]"Lean ICT – Toward Digital Simplicity," The Shift Project, October 2018
[2]"Dashboard for Sustainable Autonomous Mobility", Saujot et al., April 2018



.jpg%3Fv%3D2026-06-30T09%253A31%253A20.056Z&w=3840&q=75)







