Neither the auto industry nor politicians have yet grasped a key implication of self-driving vehicle technology: it fundamentally changes the electric vehicle range problem – which is the key limiting factor for the adoption of electric vehicles. Batteries are currently the biggest cost factor for electric vehicles; their production is costly; their enormous weight (a Tesla Model S electric battery with a capacity of about 85kWh weighs a little more than half a ton) also reduces the energy efficiency of electric cars.
But the ability to drive autonomously changes the way vehicles will be used. The first self-driving vehicles won’t be available to end users; they will provide mobility services in urban areas where most trips cover only a short to medium distance. Trips above 30km will be rare (or even impossible in a given, limited urban area) and the average trip length is unlikely to significantly exceed 10km. Thus – from the perspective of individual trips – fleet vehicles won’t require enormous battery sizes.
This means that the classical range problem for electric vehicles – which is currently seen as the major factor limiting adoption – is no longer relevant for fleets of self-driving taxis! The operators of such vehicles will not seek to maximize battery range but will want determine the optimal battery size for their usage patterns. This depends on the distance which fleet vehicles cover during the day, the geographic and temporal distribution of trips, the installed charging infrastructure and recharge speed. If we assume that the average speed achievable in urban traffic won’t exceed 30km per hour during peak times and that all vehicles will be busy servicing customers during the 3 hour morning and afternoon peaks (without any time to recharge) this means that these vehicles need to be fully charged for a range of at least 90km before the peaks. As the peak travel period ends and demand for transportation services drops off, vehicles that are idle can then drive themselves to high capacity quick charging stations and recharge so that the fleet as a whole returns to maximum battery capacity before the afternoon peak. Thus the optimal full-capacity battery range for these vehicles should be well below 200km, possibly closer to 100 than 200km. They will be able to provide mobility services with much smaller batteries than the electric vehicles that are currently being sold to private car owners, such as Tesla, Renault Zoe and others.
Self-driving cars change the fundamentals of mobility and we need to consider the effects very seriously, leaving aside our intuitions and projections which are so often based on current car-based mobility. If we examine this problem more closely, it becomes obvious that the assumption that all self-driving vehicles within a fleet should be equipped with batteries of the same size is also problematic: In urban centers a large percentage of customers only request very short trips. Thus some vehicles could be equipped with extremely small batteries and the fleet management system could channel only requests for short trips to these vehicles. Requests for longer trips could be steered towards other self-driving fleet vehicles which are equipped with larger batteries. Overall, we can expect that the total fleet battery size will be much smaller than the sum of battery capacity which a similar-sized number of privately owned electric vehicles would have.
No auto maker or new entrant in the self-driving car space has yet presented a car model that has been engineered for fleet use from ground up (going beyond the individual car design and applying a systems perspective on the fleet and its operating infrastructure). But when that happens, the architects will need to also consider whether these cars should be equipped with an ability for rapid, fully automated, battery swapping (which may also consist of adding/removing smaller battery extension packs on the order of 25 to 50km ranges). If this were feasible in small stations in a time frame of three minutes or less, then such fleets could come very close to the theoretical minimum in fleet total battery capacity with respect to given transportation demands. This would greatly reduce capital and operational costs for such fleets, increase their cost advantage over trips with privately owned vehicles, and minimize the energy costs and environmental impact of personal transportation.
It is time to recognize that self-driving cars fundamentally change many aspects of mobility and that they will be the catalyst for the electric mobility of the future. With self-driving vehicles, the classic range problem of electric vehicles vanishes (and is replaced with an entirely different problem of determining the optimal battery range for a regional mobility usage pattern). This also implies that we should be careful with our projections about the adoption rate of electric vehicles. Autonomous electric taxis could dramatically accelerate the diffusion of electric vehicles and rapidly increase the share of person-kilometers traveled in electric vehicles. Auto makers, politicians and environmentalists should take notice and move their focus from more efficient batteries to rethinking and redesigning urban (and long-distance) personal mobility in the age of self-driving vehicles!
P.S. Because fleet-based mobility services will also be available to people who still have their own (self-driving) cars, the battery range equation will also change for them. Range will no longer be such an important factor for buying a private car when it is clear that comfortable, ubiquitous fleet services are available in all cities (locally and at the destination) and that fleet vehicles can be used for long distance trips in those rare cases where the needed range is larger than the range of the electric vehicle that was purchased.