Volvo Will Add Lidar for ‘Eyes-Off-the-Road’ Self-Driving Cars on Highways

It’s 2020. Why can’t we binge Netflix as our cars drive us down the highway? Well, we’ve made progress, but not at the pace once promised. While some cars offer automated driving modes, you’re not to take your eyes off the road or hands from the wheel. Volvo wants to remedy that.

The company isn’t promising 100% self-driving cars in the near future. Instead, they’ll make mainstream cars that reliably drive themselves on highways—totally autonomously, no human attention needed. For a brand built on safety, and in light of autopilot accidents in recent years, it’s notable the company thinks that’s possible in the not-too-distant future.

To make it happen, Volvo said this week that it would begin adding lidar to production cars in 2022. They’ll also develop self-driving software to integrate lidar, cameras, radar, and back-up vehicle control systems. Once the software, dubbed Highway Pilot, is deemed safe, it’ll be sent out as an update to customers who opt in.

Cars With Laser Beams

While Volvo’s autopilot aspirations are notable, the fact they’re integrating lidar into cars aimed at the mainstream is perhaps even more interesting.

The most advanced self-driving cars (think Waymo) use lidar to navigate. It’s the spinning “can” you may have seen slapped on top of a self-driving Lexus or minivan. Lidar shoots lasers into the immediate environment to create a real-time, high-fidelity, 3D map of a car’s surroundings. Lidar is better quality than cameras or radar, but it’s also clunky and, historically costing upwards of $75,000 per unit, can itself be as pricey as a luxury car.

Which is why it’s mostly been used in self-driving car trials as opposed to production cars. Though Audi has offered lidar in some models, it’s been short range and low resolution, somewhat limiting capability.

Volvo’s lidar will be made by Luminar and is reportedly cheaper—under $1,000 when produced at scale—and sleeker, without sacrificing too much on critical capabilities like resolution and range. So, while Waymo cars still rock the can, Volvo will integrate lidar into its models much more seamlessly (see the image at the top of the article).

There is a catch, however.

Luminar’s lidar is fixed and can only scan 120 degrees. To get 360-degree coverage, like Waymo, Volvo would have to integrate multiple units into its cars. This would, of course, raise the cost (though maybe not too much, relative to lidar systems that run tens of thousands of dollars). In this iteration, at least, Volvo appears to believe front-facing lidar combined with cameras and radar is enough to safely automate highway driving.

Safety First

There are already autopilot systems available.

In Teslas, for example, a simple version of their autopilot software matches the car’s speed to traffic (an option that’s fairly widely available in many cars at this point) and steers within a lane. More advanced features include lane change and navigation between highway on- and off-ramps (this one’s still in beta). Critically, all this must be supervised by the driver, who isn’t technically supposed to remove their hands from the wheel. (A detail that’s been sadly underscored by multiple high-profile accidents.)

But this is where Volvo’s plan differs. “Soon, your Volvo will be able to drive autonomously on highways when the car determines it is safe to do so,” Henrik Green said in Volvo’s press release. “At that point, your Volvo takes responsibility for the driving and you can relax, take your eyes off the road and your hands off the wheel.”

Volvo says full autonomous driving will be limited in the beginning to certain areas and conditions. Likely, this means they’ll map particular stretches of road to help the software navigate. Over time, they’ll expand the areas where it’s available with software updates.

So, while today’s cars have some degree of what industry jargon labels level 2 autonomy (which requires supervision) Volvo is aiming for level 3 autonomy (unsupervised in some situations). The car would do the driving, but the driver has to be available to take control at any time. Not full automation across the board, then, but a step in that direction.

It is worth noting that Audi had similar, albeit more limited, plans for its lidar system to enable autonomous driving at low speeds on highways. But the company recently scrapped the feature due to regulatory challenges and liability concerns. If its lidar is more capable, Volvo may succeed where Audi fell short, but that success will depend on more than technology.

Even if Volvo’s plans don’t come together as planned, they may lead to wider integration of lidar in production cars, which could improve even basic driver assist and safety features.

Are We There Yet?

You’d be forgiven if you’re a little underwhelmed.

Four or five years ago, the tenor of the conversation was different. After Google’s self-driving project (now Waymo) made impressive progress early last decade, mainstream automakers were eager to jump on board. No one wanted to be left out.

This led to billion-dollar investments, headline-making partnerships, and aggressive predictions and timelines for full autonomy. Since then, however, many large automakers have recalibrated expectations and added years to timelines.

Still, more advanced self-driving trials continue in urban and suburban areas. Waymo has now driven over 20 million miles in the real world and adds another 8 million a day in simulations.

In the company’s Phoenix trial, they’ve removed safety drivers in some cars, and participants can hail a true self-driving ride. But when this will be rolled out more widely is uncertain. It could be soon-ish; it could be several years or more. In the meantime, niche uses—such as low-speed, geofenced applications and trucking, which Volvo is also developing—appear to be more imminent than mainstream roll-outs.

Which isn’t too surprising.

It’s a long slog between early dreams and widespread availability. Like all long slogs, it’s easy to question the reason one started slogging in the first place and when we’ll get there. Full autonomy in the real world is hard. There are just so many edge cases that can’t be anticipated, and earning the trust of regulators and the public at large takes time.

But self-driving tech is moving ahead and finding its way into real applications. If Volvo delivers on its promises, your car might soon shoot lasers as it drives you down the highway.

Image credit: Volvo

Jason Dorrier
Jason Dorrier
Jason is editorial director of Singularity Hub. He researched and wrote about finance and economics before moving on to science and technology. He's curious about pretty much everything, but especially loves learning about and sharing big ideas and advances in artificial intelligence, computing, robotics, biotech, neuroscience, and space.
Don't miss a trend
Get Hub delivered to your inbox