Pub. 57 2016-2017 Issue 1

33 FALL 2016 obstacles that must be overcome, we simply note that the sheer numbers of vehicles involved suggest something more like a minimum 20-25 year timeline for commercial adoption of this technology, rather than the 5-10 year timeline many have suggested. 70 While the panel discussion assumed a world with full availability and adoption of fully autonomous technology, we believe that such a world is at least a generation away, and that any adjustments to the distributionmodel needed to adapt to this new technology will have ample time to develop via experimentation and will naturally evolve over that period. 71 If andwhen the obstacles to full autonomy are overcome, there is little evidence or indication that autonomous capabilities, in and of themselves, will change anything about consumer demand for cars or the new car sales and distribution system. If the technology delivers on its promise, it could certainly have a profound effect on the dealers’ business model, but not necessarily on auto ownership or distribution. For example, dealer body shop and parts businesses would be tremendously affected due to the reduction in crashes expected to accompany autonomous technologies. The dealer service business model will also be affected as dealers will face increased costs associ- ated with training, equipment, and cybersecurity investments needed to ensure that they are able to securely and adequately repair increasingly advanced vehicles. The more advanced the vehicle and the greater the autonomous capabilities enabled by the software, the more important it is to have a highly-trained staff ready to repair the vehicle, equippedwith themost secure and sophisticated equipment. The nation’s dealers will simply evolve tomeet these newdemands for products and services, as they have many times in the past when faced with comparable technological developments. Moreover, in a world in which auto manufacturers are likely to face increased liability exposure (including, perhaps, strict liability) with respect to autonomous vehicles, it is difficult to envision a scenario where the warranty coverage for new vehicles is also not greatly increased, perhaps up to a “lifetime” warranty. That is because a manufacturer facing strict liability for autonomous vehicle accidents is likely to seek control over the repair of such vehicles to ensure that repairs, updates, and modifications are done securely and correctly. This will require even closer coordination between the dealer and the manufacturer to ensure consistency and leadershipwith respect to the requisite technological expertise and security investment. We are seeing the more forward-thinking manufacturers tak- ing that approach even today. One example of this expanded involvement is that autono- mous vehicles, as essentially “rolling computers,” will increas- ingly require ongoing software updates for a variety of reasons throughout their operational lives. Several of the panelists on panel four at theWorkshop discussed the potential for remote, “over the air” (OTA) software updates. Dealers believe that such OTA updates are likely, and hold tremendous promise, particularly for features not tied to driving functionality, such as navigation or entertainment software. However, OTA updates are not the panacea they are often held out to be for a number of reasons. First, there will always be a large portion of repairs that will need to be done physically. 72 Second, while many consumers will enjoy the convenience of OTA updates, many others will want to come to the dealership to ensure that the update is completed adequately and correctly, to get an explanation of the process, and to see a demonstration of any new or improved features. Third, and most importantly, any software update that could have any impact on the driving functionality of a vehicle should only be accomplished if it can be done securely , and it is far fromclear that this is possible today over the air. Certainly, security risks are present in any system, but it is also clear that a system requiring a secure physical touchpoint presents fewer threat vectors and is generally more secure than one that allows remote access. While perhaps less costly to undertake, OTAupdates, particu- larly thosemade to central operating systems or other firmware, are difficult and can be complicated, time consuming, and are simply less secure than controlled physical updates. Over just the past several years, the list of technology companies that have either reported security problems with OTA software updates, or have issued updates that have caused systemfailures or other faults, is lengthy. In fact, it would be easier to name the few that have not had such issues. OTA updates have not been consistently demonstrated to be capable of completion  NADA RESPONSE — CONTINUED FROM PAGE 31  NADA RESPONSE — CONTINUED ON PAGE 34 70 With approximately 258 million cars and trucks on U.S. roads and an average of 17 million new vehicles sold each year, even if 100% of all cars produced today were fully autonomous, it would take 15 years to turn over the entire U.S. fleet. With even the most aggressive predictions suggesting at least 5-10 years before widespread availability of fully autonomous vehicles, this suggests a conservative timeline of from 20 to 25 years before we have widespread adoption of fully autonomous vehicles. 71 Such change is not unheard of in the auto industry. Between 1955 and 1985, thousands of new dealerships opened to facilitate the sale of Japanese- and German-produced vehicles. Even if those cars may seem quaint and antiquated today, they were at the time revolutionary just as autonomous vehicles will be, containing myriad technologies that were then unheard of. Examples included the small displacement four cylinder engine and the four speed front wheel drive automatic. These are both ubiquitous today but were rare in the U.S. market as late as 1972. 72 For example, as discussed during panel four, the recently-publicized widespread problems with Takata airbags, the GM ignition switch, and the VW emissions retrofit all require physical repairs. TR. IV; 16. So would braking and powertrain systems, just to name a few.