Waymo Goes Fully Driverless

Waymo, a subsidiary of Google’s holding company Alphabet, is launching a fully driverless taxi service in the suburbs of Phoenix, Arizona.

Tempe and Mesa residents can summon a Chrysler Pacifica in the Waymo app and get driven anywhere in the pre-established 50 square mile radius.

They won’t get there quickly (the vehicle maxes out at 25 mph), but the legal team at Google are pretty sure that they’ll get there safely.

Safety was the initial promise of the self-driving vehicle hype cycle that started sometime around 2015 when Uber hired all the professors away from CMU's National Robotics Engineering Center. Eventually, machine learning algorithms would process enough data to make better decisions than humans behind the wheel.

Optimism exploded. Technologists preached about how omnidirectional sensors, fleet-level learning (where one mistake led to improvements for all), and the elimination of drunk-driving & fatigue-related accidents would save lives.

Five years later, we will actually start to learn how removing humans from behind the wheel will start to change our habits, culture, and city design.

Before we jump into some predictions, let’s clear up why Waymo is first to launch full autonomy. 

Google Has More Data

In every facet of life, Google is world class at absorbing and processing staggering amounts of data.

In the autonomous vehicle realm, that means miles and miles of open road. Only one company comes remotely close to the amount of miles Waymo has logged.

This isn’t cheap. 

Cruise was bought by General Motors in 2016. Ford & Volkswagen dropped billions on Argo Ai.

Zoox was bought by Amazon this summer.

Nuro Total was started by one of Waymo’s founding engineers and got $1 billion from Softbank.

Pretenders, like Tesla’s “Full Self Driving”, will still require drivers to keep their hands on the steering wheel and be ready to take control at all times. If the drivers stops monitoring the car, occupants face significant risk

Eleven years and billions of dollars worth of resources have allowed Google to separate from the pack.

Google’s Business Model 

Google makes more than 70% of their revenue from ads. This is both the reason they’re squarely in the crosshairs of antitrust legislation and why they’ve been running experiments for more than a decade to try and diversify their revenue streams.

Despite selling phones, FitBits, and smart speakers, hardware sales top out around $2-3 billion. Only a few percent of the bottom line.

Despite growing 53% year-over-year, Google’s cloud computing business only does $10 million in annual revenue, good for 6% of the total pie. 

Could Waymo be the business that finally tips the scales?

Probably not. In fact, it’s likely that the highest margin element of a successful autonomous taxi service would be related to a ride hailing service.

Uber did 6.9 billion trips in 2019, which accounts for ~70% market share. That was good for ~$50 billion in revenue. Unfortunately, the rideshare market’s growth was slowing before the pandemic, only about 20% per year.

Waymo has the uphill battle of building consumer trust in driverless vehicles, gaining market awareness against Lyft & Uber. Their best weapon will be to depress their prices, made possible by the removal of a human driver wage and the deep pockets of their holding company.

However, rideshare consumers have proven to be highly price sensitive, with little loyalty between the different service providers. So, Waymo’s prices will likely never rise significantly, even as the space consolidates.

Instead, Google will be best equipped to monetize via advertising, based on the ability to track user’s physical patterns in addition to their digital habits.

Google Maps is already growing into an advertising juggernaut. It will only get stronger with data built around how many people you travel with, where you physically arrive, and how long you stay there.

Surveillance capitalism? Yup.

Inexpensive transportation, subsidized by advertising, that millions of people will choose to use? No doubt about it.

Throughout history, people have always been willing to commute for about 30 minutes, one way, to and from their homes.

Cesare Marchetti, an Italian physicist, described this idea back in 1994 and it has come to be known as the Marchetti Constant.

Today in America, the average one-way commute time is about 26 minutes. 

Will that change if autonomous vehicles become pervasive?

The answer will impact trillions of dollars worth of real estate.

If the answer is yes, and commuters will put up with longer commutes if they can watch Netflix or pay full attention to other activities, then urban sprawl will explode. 

Proximity becomes even less significant in a world where the all the best restaurants (and independent chefs) deliver meals. Millions of people will head for (literally) greener pastures.

Conclusions

I’m not technically sophisticated enough to know which team has the best technology. I do know that while Amazon and Apple can match Google for resources, they lag behind on data.

Many of the best engineers that helped start Waymo have since left to start their own companies. However, autonomous vehicles are no longer a niche industry, they’re big business. Every year, technical universities like CMU and Stanford spit out more engineers who’ve been taught by the previous generation.

Google remains capable of overpaying to stay competitive. While I don’t trust them with my personal data and use ad blocking web browsers, Google leads the pack as a brand with the technical acumen to crack this nut.

When Waymo comes to Pittsburgh, I plan to take the ride.