What Does a LiDAR System “See”?

“LiDAR” is “light detection and ranging.” It is a key element to advanced driver assistance systems (ADAS) as well as autonomous driving systems

By Gary S. Vasilash

In case you ever wondered about what a LiDAR system might “see,” here is a picture of a LiDAR point cloud that includes HD resolution and long-distance range.

AEye LiDAR image. (Image: Continental)

It is worth noting that this image, from LiDAR developer AEye, isn’t necessarily the output of all systems.

Each LiDAR system developer essentially has its own approach to how the laser beams are sent out and returned and for how far and how often etc., etc., etc.

Supplier Continental has been working with AEye for the past 10 months and has now announced that it will be using the LiDAR system in its suite of hardware and software for autonomous driving systems (Level 2+ to Level 4).

In industry parlance this is known as a “stack.”

So the AEye LiDAR hardware and software will be part of the Continental stack, which will also include cameras and radar.

The rule in developing autonomous driving capability is that the more sensors the better.

Or, as Frank Petznick, head of Continental’s ADAS Business Unit, put it, “Reliable and safe automated and autonomous driving functions will not be feasible without bringing the strengths of all sensor technologies together.”

Let’s face it: If you’re in a vehicle that is driving itself, you want as many sensors and computers as possible to keep you safe when going from A to B.

Continental anticipates that it will start producing the AEye long-range LiDAR systems in production volumes by 2024.

One more interesting thing about this. While the AEye system is long range (1,000 meters), Continental has a solid-state short-range high resolution 3D flash LiDAR that it will be bringing to the market later this year.

When you hear a certain individual talking about how autonomous driving can be done with a single type of sensor and a smart processor, realize that companies like Continental aren’t creating stacks because they just like complexity.

They don’t.

They do like to assure safety.

Hyundai Launching RoboShuttle Pilot Program

Level 4 capability is rolling in the streets of Sejong Smart City. . .

By Gary S. Vasilash

Hyundai Motor Company—the parent organization of the firm that brings us Elantras and Tucsons and so on—announced it is launching a “RoboShuttle,” which is a Hyundai H350, light commercial, four-door van that is loaded with autonomous driving and artificial intelligence such that it will be able to transport people “with minimal intervention from a safety driver.”

Almost-autonomous vehicle. (Image: Hyundai)

More simply: this is going to be the basis of a ride-hailing service (there is the Shucle app developed by AIRS Company, which happens to be an AI research lab owned and operated by Hyundai). The system will determine the most-optimal routes based on demand.

The RoboShuttle will operate along a 6.1-km route—in Sejong Smart City, South Korea.

Think about that: “Sejong Smart City.”

It sounds exactly like the sort of place where there should be autonomous shuttles hailed by the Shucle app.

In the U.S. it is difficult to get the potholes filled in in many cities, and the South Koreans have established a city that is described as being smart.

Automating Big Rigs

There’s a lot of weight being hauled by one of those things. And a whole lot of processing for autonomy

Here’s something to think about the next time you’re rolling down the highway in your compact crossover:

One of those big rigs that is on the road with you can weigh 80,000 with a full trailer.

It doesn’t take a physicist to calculate that consequently stopping and maneuvering is going to require more time than the vehicle you’re in.

Plus system (look at the top of the cab) uses lidar, radar and cameras. (Image: Plus)

As drivers of those trucks tend to be on long-distance routes, developing autonomous driving capability for them is a growing area of interest.

One such company in this space is Plus, which is developing self-driving truck tech. According to Hao Zheng, CTO and co-founder of the company, they have more than 10,000 pre-orders for its system.

Here is a number from him that is even more astonishing than the aforementioned 80,000 and 10,000—even more than 80,000 times 10,000: “Enormous computing power is needed to process the trillions of operations that our autonomous driving system runs every fraction of a second.”

Trillions of operations every fraction of a second?

Plus has opted to develop its system using the NVIDIA Orin, which, according to NVIDIA, can deliver 254 trillion operations per second.

Evidently enough.

Still, driver or no, you’ve got to show those vehicles some respect.–gsv

Baidu, Geely and a Sensible Approach

Baidu is somewhat like Google, inasmuch as it operates a search engine, by far the leading search engine in China. But there are other services as well, including maps (Google), an encyclopedia (Wikipedia) and cloud storage (AWS).

So it is fair to simply describe it as a significant tech company.

Like other tech companies, it is expanding its operations. And so it should come as no surprise that it is moving into automotive.

But it isn’t like the company just discovered the space. It has been operating Baidu USA since 2011 and has been conducting autonomous driving operations in Silicon Valley for more than five years.

In 2017 Baidu announced Apollo, the autonomous driving platform that it garnered an array of partners to participate with in on the development, partners ranging from Intel to Toyota.

It is running an autonomous taxi service in a few cities in China.

Geely SEA electric vehicle platform: EVs for everyone! (Image: Geely)

Geely Holding–parent company of brands including Volvo Cars, Lynk & Co and LEVC, and lead shareholder in Geely Auto, Proton and Lotus—and Baidu have announced the creation of a partnership for the development of highly automated electric vehicles.

Geely is going to be providing the platform—the Sustainable Experience Architecture, which it announced in September 2020 as an “open source” electric vehicle platform that it would offer to other global OEMs—and Baidu the digital horsepower.

Manufacturing vehicles is a different kind of hard than the challenges associated with developing AI systems.

It makes absolute sense that a digital company would partner with a hardware manufacturer—in this case, the hardware being a vehicle, not a smartphone.

In a market where there are some 21-million passenger vehicles sold per year, where there is a comparatively low penetration rate of vehicle ownership (on the order of 173 vehicles per 1,000 people, compared with 837 in the U.S.), even a small slice of the market is still damned large.

And neither Geely nor Baidu seems to be focused on the small.–gsv