AUTONOMOUS DRIVING Who will pay for automating vehicles—automakers or consumers?
While almost everyone agrees that self-driving cars will produce enormous safety and efficiency gains, the party that will bear the greatest costs associated with adding autonomous driving technology is less than clear.
The long-term goal of automating motor vehicles is a noble one, with accidents in a future dominated by self-driving cars predicted to be reduced by as much as 90% or more. By certain calculations, this would result in an operational safety increase of at least 439% per individual vehicle.
Cost savings just from the installation of present-day Society of Automotive Engineers (SAE) Level 2-rated advanced driver assistance systems (ADAS) equate to $16,307 per vehicle over a 20-year lifespan, according to a 2015 study by the Boston Consulting Group. When the avoidance of diminished quality of life due to accidents is factored in, the savings can average as much as 98% of a vehicle’s purchase price.
With such obvious and copious benefits, one would think that the addition of autonomous vehicle (AV) technology to every newly manufactured car would be a foregone conclusion. But so far—to the extent that they’re practical options—such systems tend to only be built into or planned for higher-end vehicles, typically for reasons of expense.
SAE Level 2 ADAS can be broken down into three categories—features that aid drivers (rear-mounted cameras for parking and reversing, night vision, surround-view systems, and adaptive front headlights), features that warn drivers (parking distance control [PDC], forward collision warning [FCW], lane departure warning [LDW], and blind-spot detection [BSD]), and features that assist drivers in performing car navigation (automatic parking assist [APA], lane centering [LC], pedestrian crash avoidance mitigation [PCAM], adaptive cruise control [ACC], and traffic jam assist [TJA]). Features not considered part of SAE Level 2 (but sometimes are rated “SAE Level 2+”) include driver monitoring (DM), lane changing (LC), and overtaking.
While the cost of some basic SAE Level 2 ADASes could potentially reach under USD$1000 by the end of 2020 (as parts vendor ZF has promised), it’s predicted that SAE Level 3 technology costs per vehicle will average USD$15,361 in 2025, while SAE Level 5 technology costs are predicted to average USD$38,297 per vehicle that year (falling to USD$12,987 and USD$19,543, respectively, by 2030).
A 2018 survey of automaker engineers found that the cost of AV platforms is one of the biggest concerns around the technology, with some 33% of respondents citing expense as the most important criteria of a platform—more critical than speed, weight, size, efficiency, or power.
Consumers are generally not willing to lay out much extra money for AV features, even if they provide substantially more safety. Consulting firm Deloitte reported in January 2020 that a global poll of 35,000 potential vehicle purchasers found that in the U.S., 58% of respondents were unwilling to pay more than USD$500 for SAE Level 4 or 5 AV technology. And at least 34% of U.S. respondents said they weren’t willing to pay for AV technology at all. In addition, 31% of Americans stated they would be reluctant to pay more for a car with vehicle-to-vehicle [V2V] or vehicle-to-everything [V2X] communication capabilities—features that might be necessary to enable mass automation of all cars on a highway, for instance.
Perhaps contributing to these responses are the findings that up to 48% of Americans still have doubts about the safety of AVs, while in Europe, that number hovers around 40%. China is one of the places where consumers feel least anxious about autonomous driving technology, with 65% of potential car purchasers believing that AVs are operationally safe.
Mandates for AV feature inclusion
In some countries, governments are mandating the inclusion of specific AV features. In other cases, there may be no requirement for AV features to be integrated into new vehicles, but for cars to achieve an admirable New Car Assessment Program (NCAP) rating (administered by EU, U.S., or Chinese government-backed consortiums), certain features must be included.
Currently, European NCAP requirements state that a five-star/maximum NCAP rating is only possible if a car has a qualifying ADAS built into it. In the U.S., the National Highway Transportation Safety Authority (NHTSA) required that all new passenger vehicles weighing less than 10,000 pounds would be obligated to have rear-visibility ADAS features by May 2018.
Separate from these obligations, 20 automakers representing 99% of American auto market sales stated that they would voluntarily integrate automatic emergency braking (AEB) into their new vehicles by September 2022.
In 2017, the Chinese government stated that one of its goals going forward would be for China to lead the world in AV technology. As part of this objective, in 2018, the Chinese NCAP (C-NCAP) made AEB one of its criteria for NCAP ratings.
Obviously, these mandates necessitate an extra cost for each vehicle. While no government presently subsidizes AV technology installation in vehicles per se, a strong case can be made for doing so in order to accelerate a switchover to shared mobility solutions in the future, versus individual ownership of vehicles (shared mobility lowers infrastructure and other costs for governments). Nonetheless, Reuters has reported that presently, less than 20 percent of all cars sold come with ADAS features.
Some subsidization of costs
The average cost of a new midsize, four-passenger vehicle prior to the coronavirus pandemic was approximately USD$35,600 in the United States, USD$34,100 in Europe, USD$25,400 in Japan, and USD$26,715 in China.
Until recently in the United States, purchasers of many new energy vehicles (NEVs) such as electric cars—from, say, manufacturers such as Tesla—were partially subsidized in the form of a tax credit (typically USD$7,500) for consumers from the federal government. While officially, this subsidy was due to the alternative power source of the vehicle, the fact that AV features of the car were included cannot be overlooked; many, if not most, NEVs also come with AV features these days (in the case of Tesla, the company’s SAE Level 2 Autopilot AV platform is built into every car Tesla sells).
In Tesla’s case, the subsidy for purchasers ended in 2020, so now, Tesla itself gives a discount of $2,000 per vehicle to incentivize buyers. Of course, this means less profit for Tesla and does not in any way reduce the cost of including the company’s Autopilot.
Insurance companies have yet to offer discounts on premiums for buyers of vehicles with AV features, citing a lack of data that shows safety benefits. Some firms, however, admit that ADAS can reduce the frequency of accidents by up to 25%. Industry observers claim that the reticence of firms to offer insurance premium discounts has more to do with potential losses of revenue as auto insurance is insurers’ largest source of liquidity. A few automakers, such as General Motors, have teams working with insurance companies to provide desired ADAS data.
As of 2015, there were more than 230 million light vehicles in operation in the U.S., each with an average age of 11.4 years. While newer vehicles typically come with more ADAS features, there is an active aftermarket for installing ADAS in older vehicles, in which the consumer usually pays for the installation. Demand for certain additive features that aid or warn (rear cameras or parking assistance systems, for instance) is high due to their relatively low expense and ease of installation, whereas other features, such as night vision, are less in demand due to high cost.
Until recently, the price of installation of assist-type features—those that controlled braking, acceleration, and steering—was substantially higher (typically in the thousands of dollars), so these kinds of features were usually not added to older vehicles. However, recently, companies like Comma.ai have started to produce systems such as the aftermarket Comma Two device that permit such operations for lower-end car models at a significantly reduced expense (under USD$1000).
Servicing of ADAS
The auto service market is worth some $600 billion per year in the U.S. and more than double that amount globally. Post-purchase service for ADAS is an issue; some carmaker OEMs currently insist that customers only repair their ADAS via authorized dealerships (using genuine parts) at the risk of voiding their warranty. “The auto industry is creating the false narrative that you can either have safety, or you can have repairability,” notes Paul McCarthy, president of the Motor and Equipment Manufacturers Association (MEMA) of independent parts suppliers.
While AV technology usually results in lower crash rates, when crashes do happen, repair costs can be higher—adding up to USD$3,000 on average to a repair bill, according to service organization AAA—due to some AV equipment (such as sensors) being located on the exterior of a vehicle.
AV parts that get replaced also may have to be tested and/or calibrated. These costs can be an additional deterrent to paying for AV features upfront and/or maintaining them if they get damaged.
The Chicken-and-Egg Problem
It’s difficult to say if auto buyers’ thoughts about spending extra for AV features will change in the future. Certainly, if it could be verifiably shown that AV features reduce accidents, this could be a selling point. But, at least for SAE Level 3 and above functionality, it may be difficult to reduce accidents on a mass scale—even on highways—until a majority of cars have such technology installed. This leads to a classic chicken-and-egg scenario: why pay for the extra AV features if safety can’t really be improved until everyone has them?
While there’s been some discussion of creating separate lanes on highways for SAE Level 3-, 4- or 5-enabled cars, this is likely years away, and it’s difficult to say with any authority which states would implement this and how it would work. It’s probable that intelligent transportation system (ITS) and/or V2X technology could be used to allow and maintain access, but without any concrete examples to assess, any hypothesizing about how such a concept would be realized is mere speculation.
As long as cars can still be sold without AV features, mass-adoption of such technology may be impossible to achieve. This is likely a case where regulatory bodies may have to step in and mandate more AV features, as is starting to happen now, but to a greater extent.
If car prices rise, it’s possible that fewer cars might be sold to individuals, but this could be partially offset by the increasing prevalence of shared mobility solutions where payment for transport would follow a different model and technological expenses would not be borne by any individual passenger. It’s also likely that if more AV features are mandated for individually owned vehicles, buyers may simply shrug their shoulders and figure there’s no way to avoid a higher cost of vehicle ownership.
Indeed, if SAE Level 3, 4, or 5 systems were mandated for all cars sold in any one world region, unit costs per installed AV platform would be certain to fall markedly. Component costs of AV systems continue to drop, and even LiDAR sensors that several years ago might have cost more than USD$10,000 (or their equivalent) may cost less than USD$1,000 soon. In fact, it seems realistic to speculate that given enough volume of production, wholesale unit cost of SAE Level 3-category systems might fall well under USD$10,000 per vehicle. While that’s not nothing, for the average car buyer willing to stretch a USD$35,000 purchase cost over 72 monthly payments, the additional expense may be bearable. Of course, SAE Level 4- and 5-category systems may be a different story, at least for some time.
It’s also possible that car manufacturers might be able to lobby for short-term subsidies from federal or state governments for building such technology into all their vehicles, but it’s hard to see any such incentives becoming permanent.
For now, the most likely probability is that prices of AV platforms will keep dropping, and the models of vehicles they’ll get built into will go lower- and lower-end. Carmakers will try to outdo each other in terms of making advanced AV features standard, and eventually, an expectation of a minimum SAE Level functionality can be expected.