Autonomous transport still seems like science fiction: we know it is bound to come eventually, yet it is difficult to imagine in our daily lives. But self-driving cars are becoming a reality rapidly and this month Alphabet (Google) took us a step closer to that future. Waymo, Alphabet's autonomous car division, launched a fully self-driving car service in Phoenix, Arizona. For about the same price as an Uber, you can log-on to the Waymo One app and get picked up by a self-driving Chrysler Pacifica Hybrid minivan.
Waymo has a substantial lead in autonomous driving. It focused on full autonomy from the beginning, more than five years ahead of the auto industry. Waymo has driven over 10 million miles on public roads - the equivalent of about 10 lifetimes of human driving experience - and 7 billion miles in simulation. General Motors, the owner of Cruise, has logged far fewer miles but is the nearest competitor with plans to launch a commercial service in 2019. No one else is planning a launch until the early 2020s. Uber has fallen further behind after a fatal accident in March forced it to suspend public testing, and there are question marks over Tesla's camera-centric (vs laser scanner) technology.
There are multiple ways for Waymo to monetise its technical lead. Ride sharing is the most obvious business model with autonomous vehicles estimated to be, at most, half the cost per mile of a human operated vehicle. When autonomous safety is proven and regulation catches up there will be no job for an Uber driver - and perhaps no reason to even own a car. Alphabet is also in the enviable position to find customers cheaply through its ownership of Google Maps and the Android phone operating system. This ride sharing opportunity alone is estimated to be worth $80 billion which isn't much of a stretch compared to Uber's upcoming IPO value targeted at $120 billion.
Waymo's second opportunity is logistics: autonomous trucking will become the default, and many last mile deliveries will also be autonomous. Freight transportation is a $3 trillion market globally, and this opportunity could be even more valuable than ride sharing. Third is licensing the autonomous technology to car manufacturers, a smaller but high margin opportunity. And finally, Alphabet is in the prime position to control our entertainment and advertising for those extra ‘driverless’ hours.
Click the picture below to see Waymo in action (this is a 360 degree video which allows you to look in any direction)
Waymo could be worth $200 billion or more - close to the value of Toyota and multiple times any other car manufacturer - potentially making it the world's most valuable car company. Just like the personal computer industry, the future value of the car industry will be in the software not the metal. Waymo is one of several earnings drivers for Alphabet that we believe is under-appreciated by the market.
Comments
Jovan Vucetic
Dismissing Tesla too easily- debatable whether Waymo has any lead
Lachlan MacGregor
Hi Jovan, I agree, I don't think we should dismiss Tesla. In particular they potentially have a strong advantage in autonomous by collecting data from their existing customer fleet. My comment was perhaps too flippant as I needed to keep the article short. However many in the industry are skeptical about their approach to autonomous driving which is different to the rest of the industry. The key difference is they are opting not to use LIDAR sensors, which are sensors which use laser signals in a similar way to radar to produce a very accurate and precise three dimensional image of objects. Musk argues it isn't necessary to use LIDAR, which are currently expensive, and says they are "a crutch" to the rest of the industry. Instead Teslas are equipped with eight cameras supported by a forward facing radar as the primary technology, and 12 ultrasonic sensors for 360 degree vision. Waymo on the other hand uses three different types of LIDAR sensors, five radar sensors and eight cameras. The supporters of Tesla's strategy say that cameras combined with the advancements in machine learning will be sufficient - after all, that is what human's use. If Musk is right then Tesla should be able to develop a cheaper autonomous system and would have an advantage (though I note that Google is focused on reducing the cost of LIDAR and has reduced the cost by 90% already). But Tesla's is a higher risk strategy and may compromise their path and timing to full autonomous driving. The advantage of being first is likely to be huge given both the ride-share business model (natural local oligopolies) and the scale advantages of cost and autonomous learning opportunities. Tesla's skeptics say that the decision not to use LIDAR was forced on Tesla - they wanted to equip their existing fleet with sensors to collect data, yet couldn't afford to put LIDAR on vehicles. Some believe that Tesla know LIDAR will be necessary and will add this down the track. And coming back to the first point I made about Tesla's data. That data is potentially Tesla's real advantage as it has a fleet of tens of thousands of cars collecting data. Some argue that this data is unmatched by other makers, but I would be careful of direct comparisons. Tesla's data isn't of the same quality - they have access to the cars speed, acceleration and braking as well as short video clips. But this data is anonymous (not connected to a car), probably not contiguous and likely very difficult to train a system on. Waymo's data is far more detailed, using LIDAR and collecting the full suite of contiguous data about cars and their situations. Waymo then build full computer models and run simulations with thousands of scenarios. Don't take the simulated miles as a disadvantage vs the real thing - but rather as an important step to test the full range of scenarios. In summary, I'm certainly not dismissing Tesla and they are making great progress on a limited budget. They are doing much better than most automakers, whilst simultaneously disrupting the car industry with electric vehicles. Incredible. But I am skeptical of some of their statements which generate a lot of PR but are often not robust - like Musk's boast in 2016 that a Tesla would drive autonomously from coast-to-coast in the US by the end of 2017 (it still hasn't happened). The rest of the industry are consciously taking a different approach, and I can see why Tesla need to have taken their approach. I'm keen to hear if you see it differently though and why you think that Waymo maybe doesn't have a lead?