fbpx

Bright Machines Stock for Software-Defined Manufacturing

Traditional industrial robots do one job really well but with little or no flexibility to do anything else – sort of like game show hosts. The name of the game today is automation, and the combination of technologies like artificial intelligence, cloud computing, big data, sensors, and the Internet of Things, among others, is accelerating the Fourth Industrial Revolution. A couple of months ago, we profiled 5 Robotic Manufacturing Companies for Smart Factories. These startups emphasize software over hardware, generally employing machine learning algorithms that help robots adapt to different manufacturing tasks. One company that has capitalized on this concept of software-defined manufacturing is Bright Machines. 

About Bright Machines

Click for company website

Founded in 2018, Bright Machines raised $179 million right out of the gate that same year. We profiled the San Francisco startup in 2020 and were impressed by the technology and value proposition. Bright Machines has developed machine-learning algorithms that drive modular microfactories for the sorts of assembly-line jobs that require fine motor skills and mental acuity beyond fitting round pegs in square holes. In fact, we wrote:

Given how experienced the leadership team is, we can safely assume they signed up for the long haul as opposed to looking for a quick exit. They’re likely to go far beyond just achieving unicorn status as they work fast to disrupt an industry worth trillions of dollars.

Nanalyze MBA

We spoke too soon. Bright Machines has jumped on the SPAC crazy train. The company announced it would merge with SCVX (SCVX), a blank check company, in a move that will add up to $435 million to their warchest. Regular readers are familiar with our general disdain with these backdoor deals into the public markets, which often come with very little information to make an informed investment decision. This one is no different, but we’ll forego the usual proselytizing and jump into the particulars of how Bright Machines proposes to disrupt manufacturing. 

Software-Defined Manufacturing

Historically, robotic manufacturing systems have been one-trick ponies, capable of producing a particular kind of widget without much wriggle room. But variety is the spice of life. Today’s manufacturing environment is feeling the pressure of increasing product variability, as well as shorter product development and product cycles, according to Dr. Alexander Verl at the University of Stuttgart in a blog post for the International Federation of Robotics. The good professor has a PhD in robotics, and he conducts research in machine vision, manufacturing, and all manner of related topics. He notes that it can cost manufacturers as much as $1.3 million per hour to stop production and reconfigure a manufacturing line. 

It’s a hardware problem in need of a software solution. Software-defined manufacturing, he writes:

“… is enabling far more flexible and faster programming not only of individual machines, but also of entire production processes. Hardware and software are decoupled; with the aim of ensuring that machines can be networked and configured quickly, and that code is re-usable across different machines.”

Dr. Alexander Verl at the University of Stuttgart

He’s even put together a nice visualization of this transition from hardware- to software-defined manufacturing, with most companies falling somewhere on the spectrum:

Automation moves from hardware oriented to software oriented manufacturing solutions.
Credit: Dr. Alexander Verl, University of Stuttgart

Sitting at the top of the chart is Bright Machines, which relies on artificial intelligence as the brains behind its modular microfactories. One advantage is that each time a new task is programmed into a robot, the company’s AI-enabled system learns that task and adds it to its library of capabilities. Over time, the system becomes more and more intelligent. This theoretically increases the overall value of Bright Machine’s solution by reducing deployment costs and maximizing flexibility and scalability. At least that’s the story that Amar Hanspal, CEO and co-founder of Bright Machines, likes to tell when presenting his company’s investor deck. 

Customers, Use Cases, and Strategy

Bright Machines has deployed 60 microfactories for its 25 manufacturing customers, spanning network and data center infrastructure, electric vehicles, consumer goods, medical devices, and batteries.

Applications for software-defined manufacturing.
Applications for software-defined manufacturing. Credit: Bright Machines

How does it work? Take a manufacturer that wants to insert a high-value memory component onto a motherboard with great accuracy in repeatability. In a normal automation process, someone would program a robotic arm to perform the insertion based on geometric positions of this board. That’s fine as long as everything remains locked in place. But stuff happens over time that could cause the motherboard to become misaligned, with the robot force fitting the component and potentially causing damage. Computer vision enables the Bright Machines robot to recognize the variance and correct. Here’s a real-world example, with ROI:

Bright Machines case study.
Credit: Bright Machines

While the company continues to “drive the engine hard on new customer acquisition[s],” it sees plenty of opportunities to expand its current businesses with the old “land and expand” strategy. The idea is that most of its customers are only at the earliest stages of software-defined manufacturing, with millions of dollars of potential automation dealmaking ahead:

Customers identified for expansion.
Credit: Bright Machines

The company’s growth strategy also relies on expanding its software installations, which drives recurring revenue at more attractive margins over time. Bright Machines also expects to roll out new software applications later this year that go beyond the assembly automation software that powers its microfactories. The automation applications will focus on quality, inspection, continuous improvement, and planning. 

To Buy or Not To Buy

It’s a pretty compelling narrative so far, but the story is incomplete because we don’t have full access to the financials, as is typical with a SPAC deal. We do know the company is not pre-revenue. It took in an estimated $34 million last year and expects revenue to hit $59 million this year, based on bookings, backlog, and an ongoing long-term contract (more on that below). Cash burn isn’t ludicrous ($40 million in 2020, $63 million in 2021), which is pretty normal for a company trying to expand rapidly. Pretty much everything else in the investor deck is pure speculation at this point.

One other thing we did learn is that Bright Machines is actually a corporate spinoff that was formerly known as Autolab AI before coming out of stealth mode in 2018. It was incubated at a company called Flex (FLEX), a $9 billion electronic contract manufacturer with headquarters in Silicon Valley and Singapore. While Flex retains a minority investment in Bright Machines, it has completely separated itself from any leadership or management role. The two companies did enter into a five-year, $350 million agreement for services ranging from hardware leasing to engineering to testing services, beginning in 2018. However, that deal has since been scaled back to about $240 million. Still, Bright Machines stands to gain $18 million in 2021 from Flex, accounting for 30% of its total projected revenues.

Near-term revenue for Bright Machines
Good luck with the fine print. Credit: Bright Machines

You can read these changes in a couple of different ways. The pessimistic version is that Flex isn’t so confident in its former startup business, after all. The other version is that Flex simply has other priorities. Profits are down (just $88 million on $24 billion in revenue in 2020) over the last two fiscal years and the company is in the midst of a major restructuring. In either case, having a single client that accounts for more than 10% of your revenues is a risk we like to avoid.

Back to Bright Machines: The story is compelling but not conclusive, because we can’t draw any conclusions from cherry-picked valuations. The company has a legitimate value proportion, but we do not have enough information to make an investment at this point. Back of the napkin math shows us they’re valued in line with other enterprise software companies. Using our simple valuation ratio – market cap / annualized revenues – we get a valuation ratio of 29 (1.6 / 0.054). Here’s how that compares to some other names we’re familiar with.

Market CapAnnualized RevenuesValuation Ratio
PALANTIR (PLTR)391.36429
C3 (AI)5.80.19630
SPLUNK (SPLK)192.986.4
ALTERYX (AYX)50.47211
UIPATH (PATH)410.60767
UNITY (U)260.9428
DOCUSIGN (DOCU)381.72422
Average27

With a valuation of 29, Bright Machines is right around the average of the above basket of enterprise software companies.

Conclusion

If the merger goes through, Bright Machines is expected to emerge with a market cap of around $1.6 billion and trade under the ticker BRTM. The current share price is trading below the standard SPAC IPO price of $10, assuming that’s a fair market value to start. A slowdown in SPAC deals, combined with the potential of more oversight from the Securities Exchange Commission, seems to have tamped down the investor enthusiasm. That’s probably a good thing.

Update 3/22/2022: The SPAC deal fell through.

Share

Leave a Reply

Your email address will not be published.

  1. Another great article. Thanks. In for 600 just under $10/share. I have followed this for sometime. Somehow when I see the non descriptive name “Bright” I draw a blank. Personally I enjoy the trend where all the speculation has gone out of SPACS. 6 days since you posted this and I just spied it. As far as I am concerned SPACs give we retail investors a chance to buy into these companies. These companies still have to prove themselves and see how the reward their new capital base but I am will to wait.

    1. Cheers for the feedback Jim. Now that the speculation has largely gone out of SPACs, doesn’t it make sense to get on board only after the deal closes? Sure, all the deals seem to be closing so far, but the first time that doesn’t happen it may get ugly.

  2. I’ve worked in manufacturing plants around the globe and with automation companies for decades. “Software defined manufacturing” is only possible if a company not only owns the automated lines and software to run them, but also owns the consumer products and manufacturing plants that they can generate automation friendly designs and environments for (For Example, Tesla). As a third party manufacturer, the automated lines produced will always have to generate revenue off margins from the savings that are produced for their end customers so there exists an incentive to use them. Each new customer will have a product with a bespoke manufacturing process which will require new hardware investments, software development, IT implementations, space and process implementation iterations to make robust – which may take years to do and not scale rapidly without an abundance of partnership and long product life-cycles. A “one size fits all” assembly line, as marketed, is difficult to adapt for universal end-to-end use cases, especially with respect to complex processes with a higher need for automation that cannot easily be re-tooled. This model may apply to smaller scale companies with simple and smaller scale processes that do not have the engineering expertise or capital to get started themselves, however this space is crowded. Larger tier customers with higher revenue opportunities and much larger volumes will already have their own engineers working on automated assembly lines, software, robotics and developing their own manufacturing IP to compliment their unique process (for example, Apple). In addition, harsh pricing competition will come from the ability to do manufacturing overseas with cheaper labor for easy-to-ship consumer electronics. Their current business model, as I see it, even with AI powering robots, is similar to a automation integration business model and based on delivery of hardware assembly lines, which contradicts the market segment they are marketing themselves to be. Combined with a relatively unfocused problem statement its no wonder they entered into a SPAC deal. Even with AI powering the robots, their growth will be ultimately pegged to the margins they generate on their hardware that will come from a percentage of their customer’s savings – which is a very meticulous and historically low profit (low valuation ratio) space to be in.

    1. This is a very insightful comment, thank you for taking the time to articulate it. The fact they chose a SPAC means we don’t get sufficient information to make an informed investment decision. As soon as they provide some proper filing documents, we should be able to tease out how successful the underlying business is. Interesting comment you make in your last sentence. Robotic process automation is said to realize massive savings. In robotics automation you’re saying the savings result in “historically low profit” which is not what we would expect. If that’s the case, why implement in the first place unless you’re going to realize some massive cost savings? We really like firms that save other firms loads of cash because that will always be an easy sell to the C-suite, even more so when there’s economic turmoil.