Adithya Sanjay: Is 2019 the year of a fully functioning autonomous car?

Sunday, February 17, 2019 - 3:06pm

-

The Michigan Daily

The hype has been real. For the past decade or so, we have seen a sudden race between automakers and software development companies to build the first driverless car. What started out as simple driver assistance software has grown to fully functional autopilots that claim “full self-driving hardware.” And arguably so, as there is a market for such technology with reports showing that human error is involved in over 90 percent of all car crashes. 2019 is already set to be a big year in the automotive industry with both Uber and Lyft expressing intention to take their companies public. Not to mention Uber’s potential $120 billion valuation (that’s more than the worth of the big three automakers combined) that has investors frothing at the mouth, eagerly awaiting what has been touted as the “most hotly anticipated IPO.” But amid all of this, will this be the year we finally see the launch of a fully functional autonomous car?

Google’s Waymo is believed to be the industry leader by far. The company was reported by California’s Department of Motorvehicles to have a disengagement once every 11,154 miles, more than double the amount of miles of the next closest competitor. Even still, there seemingly appears to be quite a few imperfections that need to be fixed in the company’s technology before it can launch commercially. Waymo made many strides in 2018, hitting a total of 8 million miles logged, and was building enthusiasm for a December launch of Waymo One, a fully driverless taxi service in the suburbs of Phoenix. However, the program ended up being abandoned with the company citing “concerns about safety” as the prime factor in their decision. Nevertheless, there are recent reports of a full-service 2019 launch, with Waymo consolidating their fleet and working toward a driverless ride-hailing service in “multiple US cities.”

That being said, other autonomous programs have arguably seen a lot more obstacles the past year. Uber’s fatal crash in March that killed pedestrian Elaine Herzberg brought the program to a grinding halt for much of the year, as the city of Phoenix swiftly revoked Uber’s permission to operate in their city. Tesla has been there twice before: once in January where a user crashed into a stopped firetruck, and again in March, when a driver died after his Tesla crashed into a concrete highway divider.

While Uber’s failure did result in a huge reduction in its program, these obstacles have otherwise posed surprisingly low resistance to the tide of autonomy that has overtaken the minds of the entire automotive industry. Tesla CEO Elon Musk said in an interview that he expects not only the “carmaker will achieve full self-driving in 2019” but also that “[he] would be very surprised if any of the [other] car companies exceeded Tesla in self-driving.”

One of those “other” car companies is General Motors Corp., a company that recently announced massive cutbacks with intentions to double down and focus on the autonomous technology. GM has taken quite a hushed approach to building its inventory, acquiring the technology startup Cruise back in 2016, and have been testing a line of nearly 200 cars over the past year or so. And many are touting GM as a force as formidable as Waymo in the automotive industry, with market research professionals saying, “GM's technology is as good as Waymo's and [that] they have the capability to develop and service those vehicles.”

Also of interest is the number of software firms that have sprouted up as a reaction to this market trend. Of particular note is the autonomous driving startup Aurora which has managed to raise $530 million in its second round of funding. The company was started by the chief technical officer of Google’s self-driving program and has since recruited the directors of autonomy at Tesla and Uber. This hotbed of talent has been a strong selling point for the group, having managed to already ink deals with Volkswagen and Hyundai.

What’s more fascinating is the move made by Amazon into the autonomous car space. The company announced an investment in Aurora earlier this month, betting on an industry that could help reduce its own costs. Amazon has actually been working on a robotic delivery system itself that could help automate some of the back-end delivery processes that the company must otherwise pay for. A partnership with Aurora may just be the key to doing that.

I think we can say with certainty at this point that autonomous cars will take over the industry in the foreseeable future. Smaller localized examples of this are stretching all the way to college campuses where things like the KiwiBot exist, a food delivery robot that patrols University of California-Berkeley. Autonomous technology is poised to make a substantial impact on our daily lives. How foreseeable that future is, however, is the main point of contention. Some even argue that we are only 30 years away from the singularity — where AI can create machines smarter than humans. If that’s going to be the case though, I fully expect the commercial launch of an autonomous car by the end of the year.

Adithya Sanjay can be reached at asanjay@umich.edu.