Despite its name, Tesla’s Full Self-Driving (FSD) software is still far short of taking over full control of moving vehicles. The $12,000 (in the United States) option currently meets the SAE requirements for Level 2 (of six) autonomous driving.

Tesla’s FDS uses cameras exclusively to give the vehicle the information it needs to perform tasks like automatic lane changing and traffic light recognition. It does not give the vehicles full control with no human driver intervention.

Tesla and Musk insist that a camera-based advanced driver-assistance system (ADAS) is superior in both capabilities and cost to alternatives that use lidar (light detection and ranging) or radar. Tesla did use radar in its first version of Autopilot, but it dropped the technology last year. Autopilot was Tesla’s first go at autonomous driving, and it included software that could steer, brake and accelerate the company’s cars.

Lidar depends on machine learning (ML) and artificial intelligence (AI) to gather and interpret sensor data quickly. Volvo, which long-ago built its reputation on automobile safety, is adopting lidar as standard equipment on its new line of EVs. At the recent CES show in Las Vegas, one lidar maker showed a video comparing a Tesla camera-only system to a lidar system mounted on a Lexus.

Here’s me driving the Tesla Model Y on the left. It, of course, doesn’t have LiDAR. On the right is the @luminartech enabled vehicle. $LAZR pic.twitter.com/FlQVlQcw8O

— Patrick Moorhead (@PatrickMoorhead) January 5, 2022

Analysts at Jefferies have initiated coverage on two companies, one of which is a camera chip supplier and the other a lidar maker. Both have market caps of around $5 billion in a total available ADAS market valued at around $19 billion by 2030. The camera market is expected to grab about $7 billion of that total while lidar nabs $12 billion. Jefferies expects two to four suppliers to float to the top in each technology.

Jefferies initiated coverage on camera supplier Ambarella Inc. (NASDAQ: AMBA) with a Buy rating and a 12-month price target of $170. At a recent price of around $131 per share, the upside potential is almost 30%. Lidar maker Luminar Technologies Inc. (NASDAQ: LAZR) also was started at Buy with a price target of $20, implying upside potential of 52.7% to a price of around $13.10 per share.

Ambarella’s system on a chip (SOC) technology made a big splash with GoPro back in 2014 and 2015. Since then, GoPro’s share price has sunk while Ambarella’s reached a new high in December. Jefferies sees the company transitioning from video processing to integrated computer vision (CV) for security cameras, consumer electronics and the auto market.

Over the longer term, Jefferies believes that “CV based cameras will play an important role in advanced ADAS (Level 2), semi-autonomous (Level 2+ & Level 3), and fully autonomous (Level 4-5) advancements.” Here is the SAE chart showing the six levels of autonomous driving.

On nine functions Jefferies identifies as necessary for Levels 4 and 5, cameras score 5 out of 5 in five functions, the same as lidar and radar combined. Lidar scores a 5 to a camera’s 1 in day/night recognition, while radar bests cameras by the same difference in performance during bad weather. Radar and lidar both score 0 in color/contrast-detection where cameras score a 5. Jefferies concludes:

[W]e believe cameras are likely to prove the primary sensor for a high percentage of advanced ADAS and autonomous systems given historical reliability, role in mapping (path sensing via traffic sign detection, road signage, etc.), and low cost implementation.

The Jefferies view on Luminar is also quite positive as the rating and price target indicate. The analysts point to Luminar’s star-studded customer lineup that includes the already-mentioned Volvo, Mercedes-Benz, Polestar, SAIC and others. Luminar also has a technology partnership with Nvidia to include Luminar’s technology in Nvidia’s self-driving platform that the larger company hopes to persuade automakers to buy as a complete package.

The primary headwind for lidar makers has been the cost of the system. Test vehicles currently use lidar systems costing $75,000 or more. Luminar is targeting a sensor price of $500 to $1,000 at scale. And scaling is expected to begin this year.

If that happens, Jefferies estimates the company’s penetration into Levels 3 and 4 of the ADAS market will rise from zero today to 3% in 2025 and 18% by 2030. Those percentages reflect volumes of 2.7 million units in 2025 and 16.6 million units in 2030. For comparison, global EV sales last year came in at around 6.4 million units. In other words, Luminar’s (and Jefferies’) ramp is aggressive. The analysts note:

[F]irst mover advantage will likely prove key. As such, we are optimistic in LAZR’s positioning given technology capabilities as well as customer & partnership announcements to-date.

In addition to the lidar sensors, Luminar also sells a software stack under the brand name Sentinel, a “white-label” product that automakers can use to put their own stamp on ADAS. This is similar to the use of Alphabet’s sales of Android that smartphone makers can then customize. Jefferies does not think that automakers are going to want to spend what it would take to develop their own software, especially after pouring billions into new assembly lines to put the hardware together.

Luminar is not projecting a profit until 2024. That is also the year that Jefferies is estimating the company’s first positive adjusted EBITDA of $59 million. In 2025, Jefferies estimates adjusted EBITDA of more than five times that total, $322 million. Margins are forecast to rise to 18% in 2024 and 41% in 2025.

The risks to these projections include more competition, uncertain adoption of lidar technology (will Tesla and Musk be proved right?) and key customer risk (Volvo accounted for 64% of 2020 revenue and the company’s top 10 customers that year accounted for 94% of revenue).

Both Ambarella and Luminar shares bounced higher at the opening bell Thursday, but both traded down at noon. Ambarella was down about 1% to $132.01, in a 52-week range of $82.59 to $227.59. Luminar was down about 3%, at $13.10 in a 52-week range of $11.52 to $40.15. The SOX index was down about 3.8%, largely due to weak reports from Intel and Lam Research.

Get Our Free Investment Newsletter

I have read, and agree to the Terms of Use

Thank you for reading! Have some feedback for us?
Contact the 24/7 Wall St. editorial team.