Quick Thoughts: Waymo Gets Way More
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Tejas Raut Dessai
FOR IMPORTANT DISCLOSURES 203.901.1633/.553.9827
psagawa@ / email@example.com
February 5, 2018
Quick Thoughts: Waymo Gets Way More
- Once again, Waymo dominated the self-reported autonomous driving statistics released by the California DMV. GM Cruise made great strides to hit 2nd place, with no one else close to the leaders
- Waymo has placed an order for “thousands” of Chrysler minivans, increasing its ~600 vehicle fleet ahead of expanding its fully autonomous robo-cab service to the full Phoenix metro area by year end
- Atlanta has joined Austin, San Francisco, and Kirkland as major test cities for Waymo. We expect the company to move to offer commercial service to these geographies after launching in Phoenix
- GOOGL’s share price does not adequately reflect the $75B+ value of Waymo and we see GOOGL significantly undervalued with 30% or more upside (http://www.ssrllc.com/publication/googl-waymo-is-worth-75b-right-now/)
The California DMV released its annual compilation of autonomous driving activity as reported by companies testing in the state. The disengagement data for 2017 reveals a sustained lead for Waymo, with just 0.18 human interventions per thousand miles of driving. Moreover, commentary from Waymo parent Alphabet suggested that even this low level of disengagement wasn’t necessary for passenger safety, rather most disengagements were initiated by engineers as a precaution during unusual circumstances where the system may have performed safely without intervention. Waymo notes that it has focused its recent testing on the most challenging circumstances, making the statistical record more impressive. We note that the total miles Waymo cars drove in California were down by almost half, as the company shifted many of its test vehicles to Arizona, where it plans to launch a fully autonomous, geo-fenced robo-cab service before year end. It is also testing in Austin TX and Kirkland WA, and recently signed an agreement with the State of Georgia to allow testing to begin in Atlanta.
Surprisingly, the second best record for California testing in 2017 belonged to GM’s Cruise Automation team. GM logged 131,676 autonomous miles on California roads during the year, up from less than 20,000 miles in 2016, and reported just 105 disengagements. This 0.8 per thousand miles rate was easily the second best, and roughly equivalent to the numbers Waymo reported two years ago, in 2015. Uber, which we regarded second in the race was not required to disclose performance after its license to test was suspended and later renewed mid-2017. Tesla does not report any autonomous testing results in California, as its current systems are categorized as driver assistance rather than true self-driving. (Exhibit 1).
GOOGL began investing in self-driving nearly a decade ago and has logged more than 4 times more autonomous miles than the next competitor – currently at a staggering 4M+ miles in 25 different cities. These miles have been driven in deliberately diverse circumstances, and particularly vexing conditions were recreated at a private testing facility to gain more data and then run thousands of times in sophisticated computer simulations achieving 2.7 Billion simulated miles in 2017 alone. In a follow-on commentary through a blog post, Waymo highlighted how their system has transitioned from development of most common driving scenarios to fine tuning the system to address highly rare occurrences on the road now that they are inching closer to eliminating mishaps entirely.
Exh 1: California Self-Reported Autonomous Driving Summary, 2017
As we noted in previous research, Waymo is ready to offer autonomous TaaS services in specific geo-fenced locations in the Phoenix metro area by end of 2018 (http://www.ssrllc.com/publication/ten-investible-things-that-we-think-will-happen-in-2018/). This was further substantiated when earlier this week, Fiat Chrysler announced that Waymo has placed an order for “thousands” of Pacifica Minivans, worth upwards of $35M, to be delivered by the end of 2018. The new vehicles will be factory equipped with Waymo’s streamlined custom LiDAR kit along with a radar, vision systems and range of other sensors and Avis has already been hired to service and maintain the fleet. This purchase is an obvious next step after the company began offering rides to Phoenix residents under a sign-up only “Early Rider” program towards the end of 2017 with an engineer in the driver’s seat monitoring the vehicle. Unlike Uber’s hasty premature order of 24,000 cars from Volvo earlier this winter, amounting to a $1.4B deal, Waymo’s order appears as a measured and calculated move to expand its fleet in time for the Phoenix launch end of 2018, before it moves outside of Arizona to offer rides in Austin, Atlanta and other cities.
Success of autonomous driving systems depends on 3 key technology elements – 1. 3D mapping 2. Predictive AI Engine 3. Autonomous Controls. While all contenders started on the same page to develop the later 2, we believe GOOGL’s moat in mapping will be a differentiator. Decades of effort in developing Google Maps will translated into a monopolistic advantage as competitors have no other service that captures geographic details as accurately and minutely. These advantages might not be evident in early testing of systems but will be over time as thousands of vehicles remain deployed over the streets on several cities left on their own to navigate. This mandates an ecosystem beyond what disengagement statistics reveal about ability to succeed in this endeavor – Google has been preparing. Additionally, once mass deployment hits AI talent will be a serious bottleneck as Google continues to hoard the best mind in AI leaving the traditional car makers very less for taking.
In our December note on Waymo (http://www.ssrllc.com/publication/googl-waymo-is-worth-75b-right-now/), we estimated the project to be likely worth over $75B+ for Google right now accounting for the size of US TaaS business only. All in, Americans spend about $931B annually on personal transportation, 64% of which is local and practical – commuting, shopping, errands, etc. – addressable by services offering on-demand rides. Google could take half of the US TaaS market by 2025 which could reach $15B with runway to grow to $121B or more by 2030. Using a price to sales valuation multiple to value Waymo’s worth akin to other unicorn startups, discounting at 14.2x multiple of projected 2025 sales back to current dollars at a 10% discount rate yields $75B valuation for Waymo. This when combined with the potential GOOGL displays in its other endeavors such as enterprise cloud computing, AI assistants replacing apps as the primary interface etc. GOOGL is on its way to booking a trillion-dollar club membership – most likely, the first to get there.