If you ask almost any person who closely follows the advancement of this technology, the answer they give is probably going to be sooner than you might currently think. I would like to share with you the reasons why I agree it's sooner rather than later.
It’s been worked on longer than you may be aware:
The first automobile automation experiments began in the 1920s. They made a lot of advancements in the 1950s. The first working autonomous test vehicle was built in the 1980s. The reasons none of these efforts produced a product was the supporting technologies like the sensors, HD cameras, laser systems, radar, and processors didn’t exist, were too expensive, or needed to mature.
That all changed in the late 1990s when researchers realized the technology existed and was ready. Too many prototypes to name have been created since. In 2009 Google began secretly building and testing their prototype. When they reached the point they were confident they actually could build a working SDC, they announced it to the world and shook the auto and technology industries fully awake.
The Technology Itself:
They aren’t reinventing the wagon wheel; they’re assembling a wagon from plans. The supporting electronics that will allow the self-driving car to become a reality didn’t need to be conceptualized or invented. All of the items that make up the self driving package already existed in some form. There have been some improvements to many of the underlying technologies, not to make self-driving possible, but to make it cheaper and more powerful.
In a nutshell, this is a programming challenge, not a technology challenge.
Self Driving =/= Fully Autonomous.
Many drivers have expressed to me they feel full autonomy, driving anywhere at full speeds, could be a decade or more away. I doubt it will be that long, but it could be.
Where I don’t agree at all is with the idea that full autonomy, driving anywhere, level 5, needs to be achieved for SDC TNCs to exist. All of the limitations of a level 3.5 vehicle can be resolved with current technology. For example, if the car has an issue getting around an obstruction, it can be remotely or locally instructed on what to do or even remotely driven. Being that they currently go around obstructions all day every day, I can’t see this as a common issue.
Level 4 is the same as 5 but without the mapping to go everywhere. It’s fully autonomous within a limited geography. Level 4 is likely in the next year, if Waymo hasn’t already achieved it, IMHO.
The effort is bigger than you might think:
Forget Uber’s self driving cars, they’re infants in the SDC world. What we have to worry about is the massive group of companies that are closing in on the SDC goal.
The auto manufacturers and technology companies who are fully engaged in bringing SDCs to market, and especially TNC SDCs, seems to be an endless list of household names. The companies you may not have heard of are many times that number. This may very well be the largest effort towards a common goal ever undertaken by corporations. The worldwide scope and expertise of this group and their financial wherewithal almost can’t be exaggerated.
Companies like GM have announced to their shareholders that their entire future automobile business model will be based on SDCs and specifically TNCs. That’s a huge announcement that shows undeniable commitment. Then they dropped $500,000,000 to get a piece of Lyft. They also announced they will introduce thousands of SDC test vehicles next year.
The worldwide annual TNC market has been estimated to be $14 trillion dollars. Three quarters of the entire US GDP. That’s a LOT of motivation to stay committed and win this race.
Access to self-driving platforms:
Google has spun off their self driving car project as Waymo. Waymo doesn’t have a technology CEO, they hired a marketing CEO. Both of these moves indicate Waymo is close to introducing their technology.
This is a worst case scenario for Uber drivers. Waymo is the clear leader and Waymo isn’t building SDCs, they are leasing the technology platform to manufacturers. This business model means any auto manufacturer who wishes could lease the technology to get to market faster. Several are already working with Google on projects.
The will of the people:
In older polls there was a lot of resistance to the idea. One early poll listed the percentage of people who would ride in an SDC as 16%. Newer polls show a much more positive attitude, as high as 75% for fully self driving and 90% for full and partial.
Even for the reluctant minority, I can’t help but think most will be quickly swayed. Safety, cost, convenience, stress relief and free time are strong drivers of behavior.
And lastly, they’re already here:
Yep. The Netherlands already has a limited SDC TNC service on live roads in traffic. Several cities in Europe are about to add them. They are slow and drive a restricted route like a bus, but they are 100% self driving with no human controls.
For these reasons and others, barring unforeseen setbacks like regulations or an extreme incident involving a SDC, I have absolutely no doubt in my mind SDCs are coming soon, as in 1-3 years. This won’t be the end for Uber drivers though, that may take several more years as other companies catch up, join in, and expand.
The first SDCs you will see in actual use will almost certainly be speed and geographically limited, small, electric, and almost all will be TNCs. Unfortunately for us, that means Urban and Metropolitan SDC services in some of our largest cities. As the technology matures, speed and geo-fencing will increase and expand.
While the initial introduction will be a drop of water in a large pond, the following tidal wave won’t be. Companies from around the world will be introducing their fleets and competing for market share as fast as they can. Our work will be pushed out to the suburbs and then slowly taken away completely. As sad as that is to a person like me who genuinely likes to Uber, I also welcome the benefits for society.
So, to answer the big question, this novice’s guess as to when human Uber driving will be encroached upon at a noticeable level is about 3-4 years from today with the large metros being the first impacted. There should be at least half a dozen companies with live fleets in many cities by 2020-2021. When it will all but replace us completely is probably about 6-7 years from today. My advice and hope is that you prepare yourself now for an income source change. I am.
In a late breaking announcement California just allowed it's first self-driving vehicle to operate carrying live passangers. This French Manufacturer Easymile's product is on a set course but does use some live roads and has no human driver or controls:
Separate names with a comma.