What’s the Way Forward for Autonomous Driving?

The path to autonomous vehicles remains unclear. Does it lie in building out a fleet of ride-hailing vehicles, or do long-haul trucks on the open road offer the faster route to market?

Written by Quinten Dol
Published on Jul. 30, 2020
What’s the Way Forward for Autonomous Driving?

One of the most anticipated Silicon Valley promises turns out to be one of the toughest to deliver: self-driving cars. Since the mid-2000s, VCs and entrepreneurs have predicted that autonomous vehicles — and with them, the end of traffic and city driving — were just around the corner.

Nowadays, pundits are making more calculated (but no less optimistic) predictions of our autonomous future. Their predictions recognize the incredible complexity involved in integrating self-driving technology with fully-functional automobiles and deploying them to dynamic streets filled with humans, cyclists and other vehicles.

As that complexity has crystallized, self-driving car startups have largely turned to auto manufacturers to provide the investments and resources necessary to put their tech on the road. But the path to complete autonomy remains unclear. Does it lie in building out a fleet of robotic taxis on behalf of the Ubers and Lyfts of the world, or do long-haul trucks on the open road offer the faster route to market?

 

Brilliant, but bored? Check out all of our local software engineering jobs here.

 

pony.ai vehicle
Image: Pony.ai

Betting On Ride Hailing

With R&D facilities in Fremont, Guangzhou and Beijing, self-driving tech startup Pony.ai has positioned itself on the industry’s cutting edge in recent years. It was the first company in China to offer self-driving car rides to the public — in Guangzhou in 2018 — and last November, it teamed up with Hyundai to roll out California’s first ever public robotaxi pilot service in Irvine.

While Hyundai and Pony.ai weren’t allowed to charge riders for the service, and a human driver was still required at the wheel, the trial was a major milestone. Founded in 2016 by Google and Baidu veterans James Peng and Tiancheng Lou, Pony.ai has been working with Toyota on autonomous vehicle testing since 2019. Earlier this year, the automotive giant invested $400 million into the four-year-old startup. 

 

I truly believe that the early forms of autonomous driving vehicles will be ride hailing services.”

In response to the coronavirus pandemic and subsequent social distancing requirements, the company pivoted to e-commerce delivery. In April the company announced it was partnering with LA e-commerce platform Yamibuy to perform contactless last-mile delivery services to users in Irvine.

In the long term, however, the company sees ride hailing as a vehicle for the future of autonomous cars.

“I truly believe that the early forms of autonomous driving vehicles will be ride hailing services,” Peng said in an interview with CNBC last year. “Personal, fully autonomous driving vehicles will probably be a little further away. By having the ride hailing vehicles on the road, that business model makes sense.”

 

aurora team
Image: Aurora

The Case for Long-Haul Trucking

An alternative school of thought in the autonomous vehicle industry has gained traction under social distancing guidelines: Long-haul trucking might be the technology’s first and most impactful use case. Proponents argue that, while ride-hailing vehicles must contend with traffic hazards like stop lights, cyclists and pedestrians, the long, straight stretches of highway driving are easier for an autonomous system to navigate.

That’s not to say safety is less of an issue out on the freeway. As self-driving startup Aurora explained in a blog outlining its investment in autonomous truck technology, when trucks confront a dangerous event, “they present a significantly more complex and consequential challenge for a heavy, fast-moving vehicle than a light, slow-moving one.”

 

If you want to get to market with a safe system quickly, you can do no better than to start in trucking.”

To meet those challenges, Aurora has integrated a lidar system first built by Montana-based Blackmore into its tech stack. Founded by veterans of self-driving divisions at Alphabet, Tesla and Uber, Aurora says Blackmore’s lidar is uniquely suited for the kind of long-range sensing required for autonomous vehicles traveling at highway speeds.

“This means the Aurora Driver can detect and track objects more quickly, at greater distances, and with greater precision than was possible with existing technology,” the company wrote.

The bottom line: Aurora believes the path to profitability for autonomous vehicle tech lies in trucking.

“If you want to get to market with a safe system quickly, you can do no better than to start in trucking,” co-founder and Chief Product Officer Sterling Anderson said in a June interview with The Information. 

 

argo ai palo alto office
Image: Argo AI

Get A Partner In Your Corner

Building, testing and iterating on autonomous vehicle technology is a famously capital-intensive process, and few startups can survive without a wealthy benefactor. The most famous examples include Alphabet’s Waymo and its partnership with Fiat Chrysler, and Cruise’s acquisition by GM. Meanwhile, Pony.ai works with Toyota and Hyundai and Aurora has partnered with both Hyundai and Fiat Chrysler.

One of the biggest announcement of its kind in the last 12 months came from Argo AI, which revealed a $2.6 billion investment from Volkswagen last July. As part of the deal, the German giant pledged to merge its own autonomous vehicle subsidiary with the Pittsburgh-based startup, which has a large engineering center in Palo Alto. Argo is also developing self-driving taxis for Ford as part of a $1 billion deal, which was announced when the company came out of stealth mode back in 2017.

 

We are an independent company that has now established relationships with two global automakers.”

The trend of autonomous vehicle startups forging partnerships with established industry figures underscores how difficult the technology is to perfect. As Karl Iagnemma, who leads a joint venture between Hyundai and a self-driving startup called Aptiv, told Wired this month: “Vehicles are these massively complex systems, and to (build self-driving cars), we need to integrate them with another very complex system and do it in a way that’s reliable and cost-optimized. It’s really, really hard.”

Argo AI Founder and CEO Bryan Salesky has said that strong partnerships are the path forward for autonomous driving, especially if the services are to scale worldwide.

“One of our strategic advantages — and the thing that makes Argo truly unique — is that we are an independent company that has now established relationships with two global automakers,” Salesky wrote as he announced the VW investment. “This allows us to work hand-in-hand to approach the design, development and manufacture of self-driving vehicles holistically. From both a safety and scalability standpoint, it’s a key strategic advantage, and we believe we are the only self-driving company with such close working relationships.”

 

Hiring Now
Sojern
AdTech • Digital Media • Marketing Tech • Software • Travel