CES 2019 vs Detroit Auto Show: How Close Is Autonomous Driving, Really? 

Depending on which convention you visited this month, you may have a radically different opinion about the readiness of autonomous cars. If you went to CES in Las Vegas, you probably think self-driving vehicles are real and just around the corner. If you opted for the Detroit Auto Show, you’d come away thinking the auto industry seems stuck in the past.

It comes as part of a wider cooling in the industry’s approach to hype. Elon Musk, who claimed in October 2016 that his firm Tesla would reach full autonomy in 2017, later said in October 2018 that it’s “extremely difficult to achieve a general solution for self-driving that works well everywhere.” Uber laid off most of its workforce dedicated to the question after one of its vehicles was involved in a tragic accident. Waymo, Google’s self-driving car project operating a taxi service in Arizona, reportedly struggles to turn left.

“The AV hype has definitely been tempered significantly over the past year as almost everyone in the industry has come to realize the challenges of actually creating a robust automation system that can work in the real world,” Sam Abuelsamid, senior analyst for Navigant Research, tells Inverse.

You wouldn’t know it at CES, though, where the hype remained in full swing. YouTuber Marques Brownlee took a ride in Yandex’s autonomous taxi, which has been operating in Moscow since last October. BMW demonstrated an autonomous motorcycle in the form of a modified R1200 GS. Bosch showed an all-electric driving pod as a concept of what future vehicles may look like:

Part of it has to do with audience. People at CES, for example, are simply used to seeing inventions or concepts that won’t be ready for several years. It’s basically expected, Shiv Patel, research analyst with ABI Research, tells Inverse.

“In general, for the auto industry CES has always been about highlighting long term technology developments that are a couple of product cycles out (5-10 years),”he says.

But while future-focused CES suggests an industry still enthralled by autonomous cars, there was still subtle shift in emphasis.

“Even at CES there was less this year from the automakers and more from the suppliers and smaller companies that are trying to sell them components to enable those technologies,” Patel says. “That is, it’s more like the rest of show that is not about cars. So that means a lot of focus on sensors, chips and this year in particular digital voice assistants like Google Assistant, Alexa and others.”

The next few years for autonomous cars may lie in focusing on “level 2” semi-autonomous systems like Tesla Autopilot, where the driver is expected to keep their eyes on the road and the computer drives in limited circumstances. This was on display at the Detroit Auto Show, where Daimler Trucks announced its Class 8 Freightliner Cascadia as the first with level 2 assisted driving. Kia’s Telluride SUV includes lidar capable of following the lanes on highways. Toyota’s flagship 2020 Surpa also contains similar features like blind spot monitoring.

“What we noticed in CES and over the course of 2018, is that OEMs and in-particular tier one and two suppliers have realized that the significant short-term opportunity lies in scaling down their complex fully-autonomous robotaxi technology to consumer vehicle applications, namely advanced ADAS [advanced driver-assistance systems] functionality ie. SAE Level 2 ‘Plus’,” Patel says. “By scaling down some of the principles used in RoboTaxi operations, i.e more sensors and increased computing power, OEMs can significantly increase the performance of their current ADAS packages.”

All this adds up to an industry that now sees the importance in taking a more measured approach. Bosch predicts one million pods similar to its design could be on the roads by 2020, a goal that seems achievable considering Optibus and others have already started operating autonomous buses on very short routes. It’s a far cry from full autonomy tomorrow, but it leaves the industry free to focus on gradually taking steps that could help drivers and public transportation users today.

“There is now a broad recognition that AVs are not going to take over the world in the near term (3-5 years) but will instead be limited to very specific use cases,” Abuelsamid says. “There is also a recognition level 5 automation (that is automation that can operate anywhere and in any weather conditions) is unlikely until at least the late 2020s and likely much later than that.”

Related video: The Volkswagen I.d. Buzz Autonomous Hippie Van