LIDAR manufacturers under the radar

[kettunen]

Lidar manufacturers have gone public in 2020-2021, and discussions about them have primarily focused on the SPAC chain because they listed through it. The technology is new, companies are mostly in their early stages, and funding has been needed for investments, at least so far. A SPAC reverse merger offers a good alternative for early-stage companies, and tech sector players attract the markets. There have been some mentions here and there in other threads, but now there’s a dedicated thread for lidars, which will hopefully provide better insights into the technology itself, but especially into the companies as investment targets.

Background on the Technology:

Lidar (an acronym for light detection and ranging) is an optical device that functions similarly to radar, operating in the visible light, near-infrared, or ultraviolet spectrum. For example, a remote sensing device measures the distance to an object by sending a laser light pulse and recording the time it takes for the reflected pulse to return from the object. Lidar is used in all kinds of distance measurement, including robotics, surveying, geology, meteorology, and seismology.

Lidar differs from radar in that its wavelengths are much shorter than the microwaves used by radar. The smallest measurable object size depends on the wavelength: thus, lidar can also measure aerosol particles and cloud particles. A cloud height meter based on this principle is called a ceilometer.

(Wikipedia: Lidar – Wikipedia)

Lidar is an acronym for “light detection and ranging.” It is sometimes referred to as “laser scanning” or “3D scanning.” The technology uses eye-safe laser beams to create a 3D representation of the surveyed environment.

(Source: What is lidar? Learn How Lidar Works | Ouster)

Applications of Lidar Radars

The biggest area of interest currently is equipping cars with lidars to enable autonomous driving or driver assistance. Most manufacturers are focusing solely or primarily on this market.

However, there are many different applications: robotics, landscape and architectural imaging, traffic control and automation, drones, etc., etc.

Lidar is also available in a smaller format; for example, Apple phones already use it. Lidar is used to assist imaging in depth sensing, but also in augmented reality (AR) applications. However, the coverage in mobile devices is significantly smaller than in devices intended for industrial use (https://www.pocket-lint.com/tablets/news/apple/151476-what-is-lidar-ipad-why-arkit-measure).

Challenges of Lidar

The technology is relatively new, with the first commercial products appearing in 2007. However, its adoption is rapidly expanding. A major obstacle has been the price; sensor costs have been in the thousands of dollars, and adoption has apparently been challenging. The technology is progressing in leaps and bounds, and Solid State sensors, for example, are entering the market.

While several new manufacturers have emerged, there is also an effort to drive down sensor prices, which naturally affects margins. Conversely, demand is expected to increase dramatically if one looks at each manufacturer’s development forecasts a few years ahead. Broader demand and lower costs will spread the technology to new applications, and thus lidar manufacturers can remain interesting as investment targets. However, simply being a device supplier is unlikely to be enough; an ecosystem, licenses, and sales of applications and expert services will also be needed.

Most Significant Companies as Investment Targets

Lidar manufacturers have not historically been listed companies, at least not those specifically specializing in lidars. In autumn 2020, however, several manufacturers emerged through SPAC companies, and these have ended up on stock exchanges. Most of these players are relatively early-stage companies, with limited revenue, and significant profits are mostly expected only in 2024-2026.

Velodyne Lidar

Digital Lidar Sensors for Automation, Drones & Robotics | Ouster | Ouster

Investor materials:
https://investors.velodynelidar.com/static-files/984153ee-1604-4b85-b810-17e227a3f22e
Latest figures Q4/2020:
https://investors.velodynelidar.com/static-files/2ca8bdc5-af80-46b9-bc51-aa6adf966fbf

The most significant and first pure-play lidar manufacturer. Has been on the market since 1983, as a lidar manufacturer since 2007.

Velodyne Lidar focuses on several different product categories, which provides business support and a moat compared to competitors. Revenue comes from multiple sources, and the company is not dependent on automotive contracts, even though they are significant.

Manufacturing and selling sensors alone is no longer profitable, which is why revenue has started to decline, and the bottom line has also accumulated losses. New products have been launched, and an ecosystem has been developed around them. Service and license sales are increasing and will form a significant part of the bottom line in the future.

Original owners David and Martha Hall had to leave the company due to some disputes. New recruitments have been made, including a couple of seemingly good players.

Ouster
https://investors.ouster.com/overview/default.aspx
Investor presentation
https://s27.q4cdn.com/377532724/files/doc_presentation/Ouster-investor-presentation_12222020.pdf

SPAC merger completed in March 2021. Future figures will still show arrangement costs, but otherwise, it is a normal listed company at this stage.

Like Velodyne Lidar, it operates in several different areas, thus having a broader business base. It is a newer company but has significant clients and partners, including NVidia and Konecranes. Revenue in 2020 was $19M, with a 2021 estimate of $34M. However, the growth curve is predicted to be quite exponential, with a 2025 estimate of $1584M, a 143% CAGR.

Aeva
https://investors.aeva.ai/
Investor presentation:

02.2021-Aeva_Investor_Presentation_vF.pdf

Automotive pure-play lidar for now, possibly entering consumer electronics. 4D lidar technology, which adds an extra dimension to measurement, accelerating and enhancing measurement accuracy according to the manufacturer. Names like Volkswagen, ZF, Denso are among its backers.

Luminar

Luminar

Investor presentation
https://investors.luminartech.com/static-files/d5fe72a1-a244-4a39-bbe8-afe5c30a98ed
Has received the highest valuation, at least based on the stock chart, among manufacturers at this stage.
Includes Volvo and Daimler. A pure-play automotive manufacturer.
Revenue 2020: $15M, 2021E $26M … 2025E $837M, aiming to be a profitable company by 2024.

Innoviz

Home

Investor presentation:

Project-Ignition-Announcement-Presentation_FinalDec11.pdf

Pure-play automotive, backed by Magna, BMW, Samsung, among others. Focuses on cost-effectiveness, but competitors are also bringing out cheaper sensors.
Revenue 2020: $5M, 2021E $9M … 2025E $581M

[timontti]

Interesting start I’ll add my own two cents right away, a lidar overview in automotive from yesterday

https://www.eetimes.com/lidar-sweepstakes-draws-15-rfqs-but-no-frontrunner/

Apparently, there are about 70-80 manufacturers and no clear favorite. The RfQ is for car models coming out in 2025.

Then it’s time to pick the winners. So far, my own interest is limited by the fierce competition, but it would definitely be good to dig deeper.

[niko_naakka]

At least a couple of companies with lidar in their plans are in my own portfolio:

Microvision, Inc.

http://www.microvision.com/consumer-lidar/

Advanced lidar & perception solutions in autonomy & mobility

WIMI Hologram Cloud Inc

https://finance.yahoo.com/news/wimi-hologram-announces-plans-develop-100000155.html

[beloved_lover]

Good picks! Those Lidar companies with an automotive focus could be to my taste. Even though Daddy Musk doesn’t believe in Lidar, I have enough faith that many car manufacturers will use it to improve safety and autonomy in vehicles. Eventually, it will be standard safety equipment even in the cheapest cars, like seatbelts and airbags.

I need to take a closer look and follow up, but as @Aston_Livingstone said, the competition in the field is already quite high. Laser measurement itself is not a new invention, but automotive applications of Lidar will be in demand in the future, so there will certainly be enough room for competition. Maybe I’ll loosen up a bit on dividend portfolio purchases and add a few of these to my long-term portfolio in hopes of multi-bagging. We’ll see…

[kettunen]

Great, @Aston_Livingstone was the one I meant to ping here. Technology is significantly related to already familiar areas, e.g., SmartEye and QT. Especially since the operating models for the automotive side have already been researched, utilizing that information will speed up and aid in further research.

An interesting article from Aston that illustrates the fragmented state of the market. My own view has been that the industry will see a lot of mergers, and some players will run out of chips in the competition. This was also highlighted in the article: “A further complication is the growing number of lidar players (estimated to be 70-80 companies) crowding the market. Mergers and acquisitions are rampant.” The sheer number of manufacturers still came as a surprise.

My own interest in these areas is focused more on things other than the automotive sector. As a disclaimer, Velodyne Lidar and Ouster are in my portfolio, which reflects my view.

For example, traffic monitoring and automation is a big activity, and sensors provide benefits compared to cameras or radars. For example, from Velodyne’s website
According to the research firm Markets and Markets, the Intelligent Transportation System (ITS) market for roadways is expected to grow from USD 17.9 billion in 2020 to USD 36.5 billion by 2025, at a CAGR of 15.3 percent. This growth can be attributed to increased pedestrian safety concerns, growing traffic congestion problems and smart city developments around the world

Another perspective is, for example, modeling buildings or otherwise confined spaces, such as tunnels, mines, etc. A sensor can be attached to, say, a drone, and the view can be imaged in 3D.

This is also a Velodyne case, Emesent has made a multi-year agreement with Velodyne

[kettunen]

Ouster, on the other hand, has made a deal with Outrider, a company that automates logistics in logistics centers.

https://investors.ouster.com/news/news-details/2021/Ouster-Signs-Strategic-Customer-Agreement-with-Distribution-Yard-Automation-Leader-Outrider/default.aspx

Zero-Emission Yard Operations & Sustainable Freight - Outrider

There are a huge number of applications once you start to grasp the possibilities of the technology and the sensors shrink and the cost level comes down. Of course, falling prices will reduce turnover and margin per sensor, as will competition. For my part, I prefer to look where not everyone else is in the same market. Ouster and Velodyne, for example, operate more broadly in different sectors.

Edit: Through Aston’s link, I found another eetimes article about Ouster. Xilinx FPGA chips are used for computation. Moving from rotating radars to VCSEL and ASIC technologies is also happening with others besides Ouster.

https://www.eetimes.com/ouster-takes-on-waymo-with-lidar-diversity/

Video link found in the article:

OS2-128 Highway Drive

[kettunen]

Let’s continue with the valuations for the manufacturers’ comparison. The following is slightly outdated, from about a month ago if I recall correctly, but it still gives an idea of the valuation differences mentioned in the opening.

Share prices have come down with the rest of the tech sector at the time of writing. The impact of rising interest rates for the years 2024-2026 will also surely start to show in the current value.

It has already been discussed in the SPAC thread, but Ouster and Velodyne are clearly at a lower valuation for some reason or another. Velodyne’s growth multiple is predicted to be smaller than others, but conversely, its historical and current revenues are on a completely different level, and it has been in business for a long time compared to others.

Ouster and Aeva have therefore gone through a merger and their current tickers are $OUST and $AEVA
Innoviz and Aeye are still in the SPAC phase, meaning they are currently found under the tickers $CGRO and $CFAC.

[Heimo_Vesa]

LiDAR chain, and not a single mention of Trimble – the number one company in laser scanning. Could the reason be that the company’s business in the overhyped traffic sector is quite marginal compared to the boring construction, infrastructure, and geo sectors? Trimble’s materials, of course, include images of self-driving tractors, and a dozen products and solutions for the autonomous driving sector. Their expertise is also world-class in GPS, in addition to laser measurement, which at least a few years back was as critical a component as the LiDAR measurement device itself, when point clouds from mobile laser scanners needed to be tied to a consistent coordinate system. That GNSS expertise and satellite tracking is also a rather essential part of autonomous driving even without LiDARs.

Trimble quite closely resembles the domestic Vaisala, which over its long history has developed unparalleled expertise and position in its own measurement sector. Vaisala also, of course, has LiDAR expertise and products for the needs of the weather sector. The only problem with Trimble is its somewhat high valuation.

@kettunen

The biggest area of interest currently is equipping cars with LiDARs to enable autonomous driving or driver assistance. Most manufacturers are therefore focusing solely or mainly on this market.

Self-driving cars certainly get the most attention, but at least Velodyne, in its own slides, sees the delivery sector selling the most sensor units over a 5-year span. Interestingly, there are only 18 projects underway in this sector, so are the projects/products much more generic (and thus perhaps more profitable)? According to Velodyne, delivery includes: “touchless delivery, B2B transport, railways, and robotic delivery.” A pure guess is that these sensors would go into warehouses for picking machines.

Velodyne has one colorful slide after another and exponential curves (the Q4 presentation has already been posted), but who knows how much of this talk will materialize into action.

The slang and charts strongly resemble Smart Eye, with the difference that there is a lot of positive discussion and analysis about Smart, while Velodyne’s Seeking Alpha headlines and articles repeatedly speak of losses of confidence. Last month, a co-founder received a reprimand from the board “due to inappropriate behavior related to the board and company processes.”

Seeking Alpha’s analysis mentions, among other things:

While Velodyne Lidar didn’t provide details on the inappropriate behavior, both executives remain on the Board of Directors. The biggest concern was that the company would need to restate financials or provide a material change to the project pipeline forecasts.

To my eye, Velodyne’s curve doesn’t look stable at all, but rather completely reckless WSB (WallStreetBets) behavior.

I also don’t particularly warm to how VLDR’s investor slides are pure sales material: First, it states in big letters that it is the “First mover with highest market share,” and in small print, the justification is its own executives’ estimates. Market leadership is touted by reporting cumulative sensors sold and revenue over 16 years.

I have VLDR in my tracking position, but based on this quick scratch of the surface, I was left with the feeling that the best thing about this company is its sexy name, which in itself should perhaps be seen as a warning sign. Perhaps someone else has made similar conclusions, which might explain this:

@kettunen

Ouster and Velodyne clearly with lower valuation for one reason or another

What else happened in the LiDAR scene in mid-November? Both VLDR, CLA, and LAZR drew a dead man’s curve, but doubled their value on the same day. Did Volkswagen/Mercedes or similar announce a move to autonomous development or something?

Is there any company in this scene like Trimble that has a slightly longer track record? I’m a bit puzzled by the consistently weak revenue and business development of the companies in this thread, when scanners have been used for years in warehouse robots and similar applications. In these high-tech fields, I prefer to trust well-known players like Trimble, or at least a company that is subject to quality tracking like Redeye or Inderes. Otherwise, it feels a bit like buying a lottery ticket.

[JukkaM]

Lidar will become more common in many applications. In cars, it will first be used in ADAS functions, but in the longer term, the road network will likely be in digital form in the cloud, with real-time data updated there from car lidars and fixed lidars installed around the road network. Initially, identified objects will be transferred, where the size of the data to be transferred is small.

There are already implementations of this in China.
It’s a challenging area from an investment perspective, as not many companies will be able to get more advanced lidars into production for years.

I have investments in Velodyne and Aeva.

Velodyne Lidar
In recent years, they have acquired many customers from various industries, developed new, more affordable products with new technology, and built capabilities for mass production. So, truly profitable business will gradually start after the investments. I expect them to achieve their first positive quarter as early as next year.

The company doesn’t appear very innovative externally, which makes me a bit skeptical about it. Hopefully, they will be able to invest in ML/neural network expertise, as sensor development requires an understanding of how lidar data will be analyzed in the future.

On the other hand, the Velarray H800 seems to be the most advanced production-ready lidar and quite affordable. I haven’t found exact specifications for the technology, other than an 8-laser Micro Lidar Array (MLA). Could it be a MEMS implementation?

Velodyne founder David Hall was reportedly ousted from the company’s leadership by the founders of Graf SPAC. Perhaps it was felt that David Hall was unable to lead the company through rapid change.
On the other hand, a quick glance at Velodyne’s management team doesn’t inspire confidence.

Aeva
VW has recently communicated that they are challenging Tesla in electric vehicles. Sensors, which VW is apparently developing with Aeva, are an integral part of this. If this happens, Aeva will be a key player within a couple of years.

Aeva’s FMCW (Frequency Modulated Continuous Wave) technology is promising, and companies known for autonomous vehicles, in particular, are investing in it. These companies understand what kind of sensor will be needed in the future.
Mobileye is starting to develop FMCW-lidar (with Intel), and Aurora is doing the same.

[StockTycoon]

In his letter last week, David Hall stated that they have a different view on which products to focus on. Velodyne’s stock price, on the other hand, has plummeted especially because Ford sold its stake earlier this year.

In the valuation of these LiDARs, there is currently a pronounced preference for EV-sector weighting. Both Velodyne and Ouster are therefore significantly undervalued compared to, for example, Luminar. If either one secures a significant EV customer, it will surely be rewarded with a price increase (Ouster is looking for, among other things, an automotive sales director).

[JukkaM]

Yep, that’s what David Hall said:
“It became quickly apparent to me that Jim Graf and Michael Dee – joint founders of the SPAC – wanted to curtail my involvement in the quality and selection of products being developed, the contracts negotiated and integrity of the Company’s business moving forward. These actions, in my view, emboldened Chief Executive Officer Anand Gopalan to disregard my views.”

That somehow suggests that David Hall, according to Jim Graf and Michael Dee, wasn’t suitable to lead the company.

[Nortti]

Is any of these companies Israeli?

[StockTycoon]

Innoviz originated there.

[kettunen]

@Tiksu

LiDAR thread, and no mention of Trimble - the number one firm in laser scanning

This was precisely the reason for opening the discussion, to bring other operators and expertise to light. Great opening about Trimble and its operations. I’m familiar with Trimble’s positioning functions and GPS/GNSS expertise, while laser scanning with LiDARs has been completely off my radar in terms of general knowledge. I hadn’t delved deeper into Trimble’s other functions before, so I’ll have to investigate further.

@JukkaM

Yep, that’s what David Hall said…

As for Velodyne, @JukkaM has already opened the matter. The board has also responded to David Hall’s allegations:
“As required under Form 8-K Item 5.02, Velodyne Lidar, Inc. (the “Company”) is filing as Exhibit 17.1 to this report a letter that was delivered from David Hall to the Company’s Board of Directors on March 9, 2021 regarding his voluntary resignation. The Company strongly believes David Hall’s letter contains a number of false accusations, misstatements and inaccuracies, and disagrees with its assertions.”

Unnecessary tug-of-war, power struggle, and focus on the wrong things in any case. However, it clearly appears that there is a disagreement about the future direction and Hall had to leave.

Good recruitment to Velodyne’s management team this week:
https://seekingalpha.com/pr/18236740-velodyne-lidar-announces-appointment-of-deborah-hersman-to-board-of-directors
Background from Waymo and NTSB, thus potential for influence, e.g., in automating traffic control.

@Tiksu

What else happened in the LiDAR scene in mid-November?

The share price developments of the mentioned companies in the early stages are significantly affected by their SPAC status, and therefore significant fluctuations are visible. In Velodyne’s case, quick profits from PIPE financiers pulled the share price down to around $12 during the merger in September-October.

In November, if I remember correctly, there were rumors about Apple’s car and the use of LiDAR, which caused the companies to take a “small” jump together. Ford’s withdrawal from Velodyne’s ownership and missing revenue forecasts for Q4 due to Corona, on the other hand, pulled the share price to its current level. The 2021 guidance was also withdrawn at that time due to the effects of Corona - now possibly also other reasons with accounting . These are summarized from memory of Velodyne’s short stock exchange history.

[ghoul]

Swedish company Terranet is developing the VoxelFlow system, which they claim has significantly lower response times than its competitors. Piloting with Mercedes, at least.

https://terranet.se/en/voxelflow/

[JukkaM]

@ghoul

Swedish Terranet is developing the VoxelFlow system,

This is very interesting. If I understood correctly, 3 cameras continuously record a 2D (x and y) stream of events (they don’t take individual pictures), a laser follows the cameras’ point, and for each point, the laser gives a distance (z), thus producing a 3D x,y,z point (voxel).
I have commented in a few discussion groups that early fusion of cameras and lidar seems to be the future, meaning these sensors are combined at the raw data level. However, an efficient algorithm is needed to process this.

[Manubet]

Do these cameras process both pixel graphics and vector graphics to the endpoint?

[kelkka]

Let’s also bring up the LiDAR scanners that Apple has included in the iPad and iPhone in recent years, which allow even consumers to create 3D models quite easily. Gone are the years when you had to download a handful of open-source programs to create one poor-quality and space-consuming photogrammetry model These models can then be printed relatively easily with a 3D printer, so as long as 3D modeling is easy for consumers, 3D printers will also fly off the shelves even more.

[JukkaM]

@Manubet

Do these cameras process both pixel graphics and vector graphics data on the end device?

If referring to the VoxelFlow mentioned above, my understanding (I couldn’t find precise specs) is that it uses an event-based camera. It doesn’t take pictures but rather sends an event for every pixel change.
If a regular camera takes pictures/video, it takes multiple images, and each image is run through a neural network.

In an event-based system, only changes are processed, making it much faster.
In VoxelFlow, the laser also identifies the distance of each camera point. This creates a point cloud with x, y, z, and the color/brightness of the point (whatever the manufacturer wants to use).

Traditionally, camera images are processed in their own neural network, and lidar data in its own. Finally, based on the results of both, it’s determined what objects were identified.
In VoxelFlow, I understand that the laser and camera data are processed in the same pipeline.

[In_Der_Esche]

Applications can be found in other areas than “just” transportation, and the technology is interesting from a high-level perspective, as you said. Terranet seems to have acquired a patent portfolio for ten years from a US developer named Gerard Smits. The money from sales will be divided with a 70% share for Terranet and 30% for the developer, if I understood correctly.

I browsed the licensed patent list, but my background knowledge/studies are not sufficient to understand what is unique about them and what would provide a competitive advantage. I am mainly wondering why the patent portfolio did not interest other players, even to build a moat, even if they would not utilize them?

Hopefully, Terranet will now succeed and the success will not remain at the hype level. If one has owned the company from the beginning, only crumbs of the original investment remain after numerous share issues and warrants.