Thursday
Jan022020

From Apple to OSRAM; From Displays to Devices: a Pre-CES Year-End Review of Smart Glasses.

This is a 2019 end-of-year overview / Pre CES 2020 roundup of the smartglasses industry. We start with Apple supplier AMS’s acquisition of OSRAM that closed in December and look at OSRAM’s assets specifically in the context of AMS as an Apple supplier. Given OSRAM’s recent breakthrough in pico-laser technology, we take a deep-dive into the near-eye optics display market, specifically regarding laser illuminated waveguides. We then explore how OSRAM’s laser-tech may already have impacted Apple’s product strategy around Smart Glasses. We then explore other business developments inside Apple leading to a push-back on their rumored product launch date.

We also look at the movement of talent within the smartglasses industry specific to those with experience using laser driven light engines. We then examine how Apple’s decisions have reshaped the opportunities in the smartglasses market, with an eye to CES 2020.

This year has been a wild ride for OSRAM. The once vaunted German lighting brand is among the world’s largest manufactures of LEDs. Historically they have competed with the likes of General Electric and Philips Electronics in light bulbs, and still hold a significant share of the market in professional stage lighting. However the once high-replacement churn of incandescent lightbulbs has been superseded by long-lasting LEDs. This has had a huge impact on the profitability of lighting companies, and unlike Philips and GE that have vastly more diverse product lines, OSRAM—once a brand owned by Siemens—was spun off in a 2013 IPO.

While still among the top global players in the market, the lower replacement schedule of LED bulbs has made it much less profitable business. As a consequence OSRAM diversified, most significantly with their state of the art opto-semiconductor factory in Kulim, Malaysia, completed about one year ago.

OSRAM’s slow response to the market resulted in a substantial decline in share price. The new Malaysian semiconductor factory meant to revitalize their business was a €370 million ($436 million) investment. Having only been up and running for a couple of months, by February they acknowledged rumors that the company was on the block, initially in talks with Bain Capital joined by Carlyle Group. As the acquisition moved slowly, a competing bid was put forward by AMS (Austria Mikro Systeme), most known for their 3D face-recognition sensor used for Apple’s Face ID biometric lock technology, launched with their flagship model iPhone X.

After the monster hit of landing the Apple iPhone Face ID sensor business, AMS itself struggled to please the market. In January, AMS received a patent for a breakthrough, a “behind the screen” sensor design which could help eliminate the iPhone “notch.” However, this did not keep AMS’ stock price from taking a 13% tumble in February upon warnings from Apple pointing “to stagnating smartphone demand and a cooling Chinese economy.” By April, AMS had a 20% rebound when they announced they would be offering their face recognition sensors to Android makers. Apple previously had an exclusivity contract that appeared to have now come to an end.

Only days before OSRAM acknowledged rumors of their potential sale, OSRAM also had a optoelectronic breakthrough with a high brightness pico-laser:

While much has been noted about OSRAM’s expansion into automotive—in both lidar navigational systems, and augmented reality windshield glass—and automotive will surely be a large market for the merged companies, for the sake of this article, we’re going to focus on those technologies that have an immediate impact on AMS’ Apple devices business in general, and future Apple smart glasses in particular.

As AMS & Bain competed to acquire OSRAM, the company began revealing the fruits of their new opto-semiconductor factory: products in R&D unveiled as they moved into production. Their new pico-lasers, particularly their high-brightness green laser (shown above), was merely the first of many new product revelations, which would be appealing to AMS’s Apple client.




The competing offers—together with Germany’s cumbersome regulatory environment—led to a drawn out bidding war that lasted much of the year. It is also reasonable to suggest that OSRAM’s CEO and Board may not have been as divided as portrayed—splitting their loyalties between bidders could have been just as much a negotiating tactic.

In the end, AMS prevailed… and by extension Apple.

During the course of the bidding war—and pertinent to both Apple & smartglasses —OSRAM also unveiled:

NEW EYE-TRACKING TECHNOLOGY
While Apple acquired SMI (SensoMotoric Instruments) back in 2017, their technology is bulkier than competitors like that of AdHawk, (which Intel invested in, the same year). OSRAM’s eye-tracking solution has achieved miniaturized scale at parity, or smaller than AdHawk, as well as manufacturing the needed optical MEMS Chip (Micro-Electro-Mechanical Systems), of the dual axis gimbal variety also employed with pico-laser driven near-eye optical system we’ll cover more again below. AdHawk has a unique technology (shown at left): predictive eye-tracking software, while OSRAM’s offering comes paired with software that counters “dizziness” sometimes experienced by VR & AR users, resulting in nausea. With OSRAM’s acquisition, AMS can now match AdHawk’s hardware… yet at scale (Apple scale). The merger will allow AMS to bring a best-in-class eye-tracking solution to their largest client and others.

MINIATURIZED INFRARED LED (FOR FACE RECOGNITION)
In the midst of the bidding war, OSRAM introduced the world’s smallest infrared LED (IRLED), specifically produced as a projector to be paired with a face recognition sensor, the kind of sensor made by AMS, for the Apple iPhone, as discussed above…

…but while AMS currently supplies Apple with the proximity light sensor for the iPhone, the companion component, the iPhone’s VCSEL Projector (that sprays the user’s face with a pattern of infrared dots) is currently supplied by Finisar. OSRAM makes a competing VCSEL Projector. AMS’ OSRAM acquisition will allow them to offer the entire projector & sensor system from a single supplier. Given the conclusion of their exclusivity arrangement with Apple earlier this year, AMS will also be able to offer the full system to Android handset makers. Further, with OSRAM having just introduced their IRLED nano-projector marketed as ideal for face recognition on smartwatches and other small wearable devices, AMS could make a complete projector & sensor system for the Apple Watch, and offer biometric security to other wearable device makers.

FOCUS ON OPTICS
As I’ve covered extensively in the past, there has been a battle for dominance between waveguides and laser-to-holographic combiner displays in low-profile (consumer viable form factor) near-eye optic display systems.

After some context I’ll show how AMS’ OSRAM acquisition fits into this contest.

APPLE’S OPTICS (TO DATE)
After investing $200M in Corning’s material optics group, the same month they poached optics engineer Michael Simmonds from BAE Systems’ waveguide team, Apple followed with their acquisition of Akonia Holographics, maker of holographic waveguide displays. Coupling Apple’s prior acquisition of MicroLED maker, LuxVue, to a MicroLED manufacturing contract awarded to TSMC of Taiwan, it appeared that Apple’s display module for a consumer smart glasses product was fully formed: a MicroLED light-engine illuminating a holographic waveguide.

HYBRIDS ARRIVE
Entering the fray this year have been laser-to-waveguide hybrids: using a pico-laser to illuminate the input coupling of a waveguide. These new hybrid displays—introduced in the past year by both DigiLens, and Creative Microsystems—have received a lot of attention, given their incorporation into Microsoft’s long awaited HoloLens 2 (with display developed in partnership with MicroVision).

The principal advantage to these new pico-laser driven waveguide designs is that they’ve been shown to produce a wider FOV (Field of View) than possible with existing MicroLED or LCD light-engine modules.

How they achieve a wider FOV however, is slightly different in each case. From Microsoft/MicroVision, to DigiLens, to Creative Microsystem, they’re all doing something a little different to achieve a wider field of view, but they’re all achieving it by illuminating their waveguide with a pico-laser rather than a panel based light-engine such as a MicroLED. This is interesting as the whole industry was leaning towards MicroLED for the future direction in waveguide based smartglasses’ light engines… then suddenly!

The pivot required a closer look (I confess, I overthought this for some time).

While everyone is doing something different, ultimately it comes down to pixel size. A high quality MicroLED can be manufactured that produces a point of light of about 5 μm (μm = micrometer or micron = one millionth of a meter). That sounds tiny (and it is), however a pico-laser can produce a point of light of about 0.5 μm (one-half of one millionth of a meter, or about 1/10 the size of the smallest mass producible MicroLED). Microsoft, and their partner MicroVision have been first to market with a laser light engine to waveguide hybrid display. Some have suggested this is over-engineered—it is a stupendous feat of optical engineering—Microsoft has shown that it can be done, and mass produced. Most of what is publicly known about HoloLens’ display comes from a talk given by Alex Kipman, two months ago at ETH Zurich. I had previously reported Microsoft partnering with MicroVision (through a contract, a license, and a source), but with no official acknowledgment from Microsoft or MicroVision, many saw further confirmation in one of Kipman’s ETH slides. Because I’ve been asked by others about Microsoft’s waveguide supplier, I’ll make a speculation: it appears to me that Holographix is their likely waveguide supplier, though this is not confirmed (nor has it been reported elsewhere). While many like MagicLeap are using nano-lithography, others like DigiLens, and Apple owned Akonia are using holographic gratings, Kipman specifically describes a stamped manufacturing process for surface relief waveguides that sounds as though he is describing the manufacturing method known to be employed by Holographix (described here in this archived version of their website). Holographix is a high volume manufacturer of custom replicated optics and optical assemblies. They supply components to defense / aerospace, telecommunications and other industries. For waveguide display systems, Holographix is best known in the trade as the supplier to the avionics arm of BAE Systems (British Aerospace) supplying waveguides for both dash mount and in-helmet (near-eye) HUDs. Holographix has the capability to meet both the volume needs and precision requirements of HoloLens 2; they make SRG waveguides of the kind used in HoloLens 2; and they have a unique manufacturing process that fits the description from Kipman’s keynote. This leads me to believe they are Microsoft’s supplier for HoloLens 2’s waveguides. That assessment has not been confirmed at the time of publishing (I will update if/when new information comes available). Given I’m often asked and Microsoft has not disclosed, that is my qualified speculation.

DigiLens has for some time been the market leader in FOV for waveguides. One of their principal stumbling blocks has been that the resolution of available light engine technology has not kept up sufficient to expand the FOV without sacrifices to display quality. If the resolution is higher, the panel may be too large for the input grating. If the input grating is modified to accommodate a larger panel, then you have a very bulky device. DigiLens has produced some larger waveguides for automotive dash-mount HUD prototypes, and commercially available avionics HUDs (the DigiLens / Rockwell Collins produced AeroHUD provides similar heads up cockpit data as the Holographix / BAE Systems’ LiteHUD) but the large light engines in these displays are not a viable solution for anything head-mounted. In December 2018 DigiLens unveiled their first pico-laser driven, stereoscopic waveguide display, a reference model with a 50° FOV. They have stated that they are working with pico-projector manufactures to make a laser projector with a sufficient resolution to scale with their FOV. The light engine remains their current limitation. Laser projectors are the solution they are currently pursuing to solve for that.

TriLite, based in Vienna, Austria, is one such laser projector light engine supplier. After 2019’s VGA resolution dev kit, they’re now taking orders for 2020’s 720p resolution Trixel 2 module. Their modules resolution and corresponding FOV growth projection is on the one hand aggressive, yet on the other shows how light modules currently lag behind waveguides at providing adequate resolution for wider FOV in smart glasses displays. We’re still a long way from “retina display” resolution:

Then we have Creative Microsystems (CMC), a somewhat obscure U.S. military contractor who supplies the Army with price-is-no-object AR headsets with a 110° FOV…



…I interviewed CMC co-founder, Bill Parker earlier this year.

As we spoke he began to describe what set their design apart, and allowed them to achieve their staggering 110° field of view. With enthusiasm he began explaining how their waveguides had specially designed input couplings. That an input coupling (a.k.a. bragg grating) for a typical waveguide was designed for “collimated” light—that is to say, rays of light that are parallel—as is the case when the light engine is a flat plane, such as a MicroLED. However, because a laser emanates from a point (typically bounced off a MEMS mirror), the light rays fan out. As they fan out, the angle of projection changes. While optics exist to collimate light, he explained how CMC’s input coupling was designed from the ground up to receive light pivoting out from a single point, rather than for a panel display. As he explained this I reached for an audio recorder app, and asked to get this on record.

Once recording, Bill seemed to dodge re-explaining how their input coupling was customized for a laser—I felt he may have revealed more of their secret-sauce than intended. He did, however, give additional detail that I will quote here:

So the wide field of view challenge, when you have a source like an LCoS panel or a DLP—those things are big—so they’re starting out with a large area that you have to illuminate uniformly… and that goes back to the understand of how you put light on there… the étendue—how efficient that light is going to be able to maintain its brightness going through an optical system—is constrained in a system that has a lens, and that lens can have an aperture, it’s going to have its exit size, and its going to have a pupil. The pupil where the least confusion of the light is, where you get the best contrast, from the pupil being the right size for the source and the étendue. So you put all that together, and a laser has an advantage because its source size is extremely small… and because of that it can get a large aperture equivalent without have to sacrifice brightness. And it can do that with the right grating [coupler] combination very efficiently. The price you pay is that with a narrow wavelength source, the interaction with the diffraction gratings—whether they’re holographic… like a DigiLens approach, or with a surface relief like Microsoft and everybody else—those bounces are constrained by the bandwidth of the light source. A laser is very narrow, an LCoS illuminated with LEDs is very broad, so you can fill in between so you don’t get as much of a screen door effect, and that is sort of the other side of the coin. So you get the potential for doing wider field of view, higher brightness, higher contrast using laser, but you get this hit in the construction of the device. So what we worked on primarily was how to get a really good image, really high contrast, very high brightness, and we pulled that off.

I inquired again, “You made a comment before about the input grating being ‘angle specific,’ so is that saying, like if you had [a panel such as MicroLED] and so it’s wider, and you’re shining directly in, whereas a laser is coming from one point… and so your input grating has to capture the angle of the laser…”

Yeah, the name of the game is total internal reflection. So you want all that light to be propagating through the input grating, and through the [waveguide], and back out again… If you think about the math of that: the steeper the angle, the less likely it is to be able to take that turn from the grating, and get in to the total internal reflection angles. That’s a challenge that is overcome with… the right fabrication and design. Everybody’s got the same physics to deal with. In our case we got to it first.

When I inquired with Bill for a follow-up, he was not available for further comment. As a consequence, I created the technical illustration above from my jotted written notes. The illustration is more conceptual than literal, but if the input coupling (or input bragg grating) is explicitly designed to receive light emanating from a point, rather than collimated—together with Bill’s further elaborations above—Creative Microsystems have shown that a much smaller laser-based light engine can drive a much wider FOV.

LASERS VS WAVEGUIDES TO HYBRIDS
Up to this past year—prior to their use as a light engine to a waveguide—lasers displays had been an alternative, competing technology to waveguides. While laser based near-eye optics display systems were first successfully commercialized by MicroVision almost two decades ago, these were bounced off of very simple beam-splitter combiners. Only more recently did a pair of companies out of Switzerland—
Composyt Light Labs, and Lemoptix—partnered to build a system that embeds holographic mirror optics into a traditional eyeglass lens, for an approach now know as a “laser to holographic combiner” display. After their acquisition by Intel, this display tech was used in the briefly debuted Intel Vaunt glasses, before being popularized in Focals by Canadian smartglasses startup, North. Intel invested in North, who then consolidated the Composyt Light Labs & Lemoptix IP with their own extensive portfolio.

THE PEOPLE
This has significant talent / human resources implications on those teams now pursuing laser to waveguide hybrids, as we know the expertise in laser illuminated near-eye optics is, at this stage, somewhat limited. Hence, following the talent can reveal a bit about who is pursuing what.

Composyt Light Labs & Lemoptix IP may have passed through Intel and landed with North, but the entire optics teams of both companies were hired by Magic Leap. While Magic Leap’s current developer device employs nano-lithographically manufactured, surface relief waveguides, their IP portfolio shows an extensive exploration of exotic display concepts. By bringing on board the core of both Composyt Light Labs & Lemoptix’s optical engineering teams, it heavily suggests that laser to waveguide hybrids are in Magic Leap’s future.

WHO IS APPLE POACHING?
Also notable are the seven known North employees—maker of Focals—who have joined Apple. These hires were not in a block, like Composyt & &Lemoptix joining Magic Leap, but rather a migration from various levels including four intern-level engineers (even one retail employee from their store in Brooklyn); but up the value chain includes Product Design Engineer, Farhan Hossain. Hossain came more as a career migration, after an earlier departure from North, even spending time at micro-display panel manufacturer eMagin. Gabriel Reyes may be the most significant poach. A research scientist with a doctorate from Georgia Tech, his dissertation on wearable computing and recipient of the Google PhD Fellowship. Reyes had been North’s chief Research Scientist and New Concepts Project Lead at North—likely as much a blow for North to lose, as for Apple to gain…

Apple’s most significant talent acquisition in optics came with their purchase of Akonia Holographics. While the acquisition was not revealed until August 2018, the deal could have been done perhaps as early as February 2017. My own sources said they were rumored to have had a field of view breakthrough, but nobody knew much. The team has remained in Colorado, and equally mute since the acquisition.

What I recently discovered is that Akonia has continued filing their patents under their own name rather than Apple’s. Because they’re assigned to Akonia, they remain undiscoverable under any Apple patent search. Most significantly—rumors proven true—in September of this year Akonia was indeed awarded a patent simply titled “Field of view enhancement.” The filing was in April of 2017, and was published in August of this year. I’m not aware that it has yet received any media attention in the trade press.

While much of the patent art is rather abstract, Fig. 3 most clearly illustrated their innovation (I’ve replaced black-and-white pattern fills with color for clarity below). My cursory evaluation is that it is reminiscent of the multilayering holographic waveguide technology developed for wide FOV by DigiLens (not implying infringement, only to say they appear to be doing something similar.). It also has some superficial similarity to a layering technique employed by Magic Leap to achieve multiple depths of field (NOTE: This is not to be confused with a different optical display patent published under Apple’s name the same month {20190285897}, that did receive some media attention, which also related to a wide field-of-view, but was based on a laser-to-holographic-combiner system more similar to that employed by North’s Focals.).

Below, we have another patent that received a bit of attention last month. Unlike Akonia’s wide FOV patent above, this one was filed in Apple’s name, and therefore quickly picked up by the likes of the Patently Apple blog.

Like many Apple patents, it is written in a style that protects the invention while keeping as many details as ambiguous, or as broad as possible. What I have sussed out from this filing is that it is written to cover (1) a waveguide near-eye optic display that uses holographic input and output couplers, and that (2) the light engine can be one of many sources, but gives special attention to LBS laser projectors. Further, (3) it may feed into an optical collimator lens, and that (4) within the waveguide, the light path includes a “light-redirecting element” between the input and output couplings that may be similar to DigiLens’ “fold grating” design. For the points above, some of the operative language is quoted here:

The display system defined… wherein the input coupler and the output coupler each include a holographic optical element… A display system… that receives light from the display unit and that redirects the light out of the optical system, wherein the optical system includes a first input coupler, a first output coupler, and a waveguide; and a light-redirecting element interposed between the input coupler and the output coupler… a display unit based on a liquid crystal display, organic light-emitting diode display, cathode ray tube, plasma display, projector display (e.g., a projector based on an array of micromirrors), liquid crystal on silicon display, or other suitable type of display… Light… may be collimated using a lens such as collimating lens…

To further decipher: a “projector display (e.g., a projector based on an array of micromirrors)” refers to a laser to MEMS Chip with micromirror(s) on an electromechanical gimbal. It also seems a bit cheeky for the patent filing to including “cathode ray tube,” seemingly implying some of these other display technologies may also be (contextually) obsolete, and highly unlikely to be used as their light engine. It is my belief that while they didn’t wish to limit themselves, they also didn’t want to tip their hand as to which direction they’re pursuing.

As with the prior Akonia patent the Apple assigned patent also addresses the FOV issue:

It can be challenging to design devices such as these. If care is not taken, some of the field of view produced by a near-eye display may not be viewable from a single eye position.

But one of the more intriguing passages is buried in the various forms that a secondary “light-redirecting element” may be configured:

In arrangements where the light-redirecting element is interposed between the display unit and the input coupler, the light-redirecting element may include a secondary input coupler and a secondary output coupler on a second waveguide.… A light-redirecting element may be used to redirect or redistribute light that would otherwise be outside of the user’s field of view towards the user’s eyes. The light-redirecting element may be interposed between the display unit and the input coupler, may be interposed between the input coupler and the output coupler, or may be integrated with the output coupler.

The patent filing goes on in this manner describing a full exploration of various input / output coupling configurations, including a completely independent secondary waveguide (which would presuppose having its own secondary light engine), within the same lens structure—or even two light engines with independent input couplers, merging into a common output—as a means of expanding the display’s the field of view.

My conclusion with Apple’s developing optics’ IP is that they are designing in a direction that is consistent with DigiLens. I have in the past cited DigiLens as a potential Apple acquisition target. I have doubled down on this in the past, and will triple down here: DigiLens and Apple are pursuing complimentary technological paths in their optics. It would be reasonable to suggest that Apple may still give DigiLens a look.

Further, circling back to Apple’s supplier AMS’ acquisition of OSRAM that opened this article: DigiLens is working on LBS as a light engine to holographic waveguides. AMS’ OSRAM acquisition has brought them into their existing stable of suppliers, just as they’ve introduced an industry-leading laser product, right at a time when LBS projectors are becoming the preferred light engine for waveguide near-eye optics. DigiLens expertise dovetails neatly into that product development strategy.

Starting around 2017, most notably with a scoop by Bloomberg Apple analyst, Mark Gurman, many industry prognosticators had pinned 2020 as the date that Apple would be introducing a consumer smartglasses product. These numbers jibed with my own sources, and until a mere couple of months ago, 2020 was still the projection date. It was even moved up: most, including myself, had been predicting a fall 2020 reveal, for a 2021 market release. Then TheInformation—tech industry paywalled news site—broke the story of an all-hands-on-deck Apple management meeting in Steve Jobs Theater, at Apple headquarters in Cupertino…

…to announce the delay, and reset of Apple’s smartglasses product. This was quite an unusual precedent for Apple, who typical silo new product development behind many layers of “need to know” internal secrecy. Given the years of hype, dare I suggest, this event seemed designed to leak. And leak it did.

This occurred after a tumultuous year for Apple, most of it swirling around their future smartglasses. In early February, Variety broke the exclusive that Avi Bar-Zeev—presumed product lead on Apple’s smartglasses—had left Apple. Then in late March, fashion editor Vanessa Friedman’s Is Apple Saying Goodbye to Fashion? ran in the New York Times. By June, Jony Ive had resigned as Chief Design Officer. But Ive’s departure had been imminent for years. The New Yorker’s February, 2015 The Shape of Things to Come, was not even the first to openly discuss Ive’s coming retirement.

The real story was not that Jony Ive’s was retiring—with extensive coverage of his new firm LoveFrom—but rather that Marc Newson was departing with him. Newson was officially hired by Apple in 2014, but his hire was kept a secret until his first product—the Apple Watch—was revealed late that year. Ive’s departure was long planned… with Newson’s hire his presumed succession plan.

Jony Ive and Marc Newson had been good friends for many years. Before bringing Newson into Apple, they worked on a charity project together—the 2013 Bono organized (RED) Auction at Sotheby’s—clearly testing how well they collaborated as designers. Designs included a one-off Leica camera, an aluminum desk, and most tellingly, a Jaeger-LeCoultre branded wristwatch.

The Jaeger-LeCoultre collaboration was not Marc Newson’s first watch. For a time, he designed an entire line of watches for Ikepod. Perhaps most significant, and least reported, is that while working on the Apple Watch, Marc Newson also designed a line of fashion eye-frames for Safilo. Any regular reader of my blog or Twitter account will recognize I’ve been using variations of Marc Newson’s Safilo eye-frames as a stand-in for Apple smartglasses for years (including the title image for this article). This cannot be overstated: Marc Newson was already designing eye-frames while working on the Apple Watch design.

The departure of Marc Newson leaving together with Jony Ive was a wholesale rebuke of Apple’s future business direction by both the current and presumed future leadership of Apple’s product design team… and intimately connected to the product direction of Apple’s future smartglasses.

Ive and Williams had rumbled in the past, this was not their first joust. Ive wanted to leverage the Apple Watch to pivot the company into a technology-imbued fashion brand. Williams wanted to position the Apple Watch as a medical and health-tracking device. While Apple pumped money into the fashion arena, landing product placement on the cover of VOGUE China in 2014, ran a twelve page advertorial spread in VOGUE (US) in 2015, and Ive made his final stand with Apple’s sponsorship of the MET Fashion Gala in early 2016… nonetheless by September 2016 Apple quietly discontinued the extravagantly priced Rose Gold Apple Watch, and in the same month publicly pivoted the wrist-worn gadget to a health-tracking product strategy.

Jony Ive lost, Jeff Williams won.

With the rather public departure of Angela Ahrendts in January 2019, it was also noted that Paul Deneve, former CEO of Yves Saint Laurent, who was presumed to lead sales on future Apple smartglasses, had discreetly departed Apple more than a year before. Patrick Pruniaux, poached from Tag Heuer to lead Apple Watch sales, had already slipped quietly out the door years ago, to run Ulysse Nardin. Every former fashion executive in Apple leadership was now departed… and soon, so would Ive & Newson.

In April 2019, an Apple patent was published showing smartglasses as a sophisticated health-tracking device. By June, Ive & Newson were out. By July, stories broke throughout the tech and business press that Tim Cook had found his successor in “heir apparent” Jeff Williams.

For months Apple kept a confident face, and rumors leaked that they may even be revealing their new smartglasses as early as second quarter 2020. By November 11th The Information broke their story of Apple splitting their AR hardware into two products: a “visor” headset in 2022, followed by smartglasses in 2023.

I owe props here to a colleague: Robert Scoble had been predicting multiple Apple devices as early as April, and by August details emerged of a “visor,” with smartglasses to come later. In spite of skeptics (often including myself), it appears Scoble’s source was accurate.

Even speaking as a “design guy” myself—and questioning Apple’s commitment to design leadership—I still find a great deal to admire in Jeff Williams.

Williams and Cook share a lot in common, and it’s easy to see how they relate. Just like Tim, Jeff came up through supply chain management, eventually rising to COO—literally Tim Cook’s identical career path, and last title under Steve Jobs before becoming CEO. This is important not just because Jeff & Tim understand each other, but because in a world fraught with geopolitical uncertainties, understanding how to navigate international supply chains, maintain high quality standards, and high-volume production without supplier disruptions… and without getting margins crushed by tariffs in a trade war. This is the new normal.

I did a cursory exploration of Apple suppliers & acquisitions, most relevant to a future smartglasses product—also relevant to other mobile devices—and it is clear that Apple is building out a supply chain less dependent on Chinese suppliers. Like many American companies, their greatest dependency being final assembly.

In the midst of a product pivot, the AMS OSRAM acquisition provides Apple with not only the opportunity to reevaluate their display strategy, but also to further insulate their supply chain from geopolitical risk. While OSRAM is German (AMS is Austrian) their new opto-semiconductor factory is located in Malaysia. Malaysia is in an excellent position to benefit from U.S. companies looking to diversify out of Chinese manufacturing dependency. Whether the current U.S. administration remain in office beyond 2020, the existing trade row has been a wakeup call for many industries and companies to reassess their supply chain, in light of exposure to geopolitical realities.

For a quick look at Apple’s core components, of any future smartglasses product—both acquires and suppliers:

…but these countries represent the nation of origin of the parent company, not always where manufacturing is done. Apple publishes an annual supplier list, of their top 200 suppliers, representing 98 of their procurement expenditures. We know OSRAM’s relevant opto-semiconductor factory is in Malaysia. U.S. based companies like Corning, listed above also manufacture in Korea, Japan, and Taiwan; but not in mainland China (and therefore not impacted by the current trade war). AMS of Austria, who acquired OSRAM, has their own manufacturing spread between Austria, Singapore, and the Philippines.

Apple principally uses China for final assembly. They recently announced their plans to keep their MacPro factory in Austin, TX, while seeking exemptions for some tariffs.

If I were advising Apple (or anyone) on smartglasses’ final assembly, I would say look to Italy. Home to most of the fashion eye-frame manufacturing industry—including Luxottica, Safilo, and others—they produce eye-frames of high build quality, “Made in Italy” is recognized throughout the world as a symbol of premium manufacturing particularly in eye-frames and accessories, and Italy enjoys favorable trade relations with the United States. Because Italy also enjoys favorable trade relations with China, some tariffs could even be avoided by moving Chinese manufactured components through Italy, provided they meet certain requirements regarding the nation of origin of the greater percentage of the end product’s value. Italy is well positioned to benefit from a newly emerging consumer smartglasses market.

By pushing back their launch target by three years, Apple effectively told the rest of the industry, “We don’t yet have it, and the rest of you now have three more years to prepare for our entrance to the market.” This changes the game. Paradoxically, Apple’s commitment to the space fuels investment, but their ultimate entrance to the space will cool investment.





I’ve spent most of this article on the bleeding edge of display technology, here we’ll explore the alternate product strategy: starting with the minimum technology, and adding features only when they are able to be incorporated within a consumer viable form factor.

For the most minimal of minimum viable products in the consumer smart glasses space, there have been a flurry of audio only smartglasses. Most notable among them include Bose, Huawei, Amazon, and most recently Kopin, each leveraging their own strength: Bose, obviously a market leading brand is speakers and headphones; Huawei focusing on the Asian market, in partnership with Gentle Monster eye-frames; Amazon using their Echo Frames to extend the penetration of their Alexa virtual assistant, and Kopin for their Whisper™ tech, industry leading noise cancellation technology. While these are all notable, for the purposes of this article, I’m going to focus on those which—however minimally—still retain a display.

Of those with a minimal display, I will focus on two optics strategies, and three companies, across two business models.



FOCALS BY NORTH
North—the Canadian company previously known as Thalmic Labs—raised an exceptional amount of investment as far back as 2016, principally from a combination of Amazon Alexa Fund & Intel Capital. Their consolidation of optics IP via Intel’s Composyt Light Labs & Lemoptix acquisitions has already been discussed in this article. For Focals version one, North chose to go with laser-to-holographic-combiner optics.

In recent weeks North has weathered a long gathering storm. They’ve put on a good face announcing the Focals version two, coming in 2020, but will not be participating in CES. I’ve written previously about North’s substantial optics IP, most notably waveguides within a prescription lens. However, they have an uphill battle. They’ve lost talent, including as mentioned above to Apple, and a recent exposé in The Verge. Product managers had expressed early concerns regarding their price-point. This alone would not be so concerning, as they responded to market realities. However one of my own criticism has been of their through-the-air-projection system’s susceptibility to obstruction, as the article noted is problematic for women who use mascara (or frankly anyone with long eyelashes, or even hair/bangs of a certain length). More troubling, North had extensive layoffs earlier this year, that resulted in the loss of $24M in funding from the Canadian government (presumably based on “job creation”).

Of further concern, North’s aggressive move to build their own brand, including North branded brick and mortar retail—a capital intensive play quite ambitious particularly for a startup of their stage—showed admirable confidence, but also came at high risk.

BOSCH
It was only a month prior to North’s announcement of discontinuing version one of Focals, that Covestro—a spin-off of Bayer MaterialScience—announced their partnership in co-developing North’s holographic combiner lenses. No sooner did Covestro make these lenses available on the market to other third parties, did Bosch announce a white-labeled smartglasses module for other eyewear brands to make what amounts to Focals knock-offs (Covestro and Bosch have pre-existing development partnerships, and while not announced, Covestro seems likely to be supplying the same holographic combiner lenses to Bosch). The Bosch white-labeled knock-off design was announced on December 10th… the very same day that North announced the discontinuation of Focals version one.

Stephen Lake, North’s co-founder and CEO is one of the more promising and ambitious entrepreneurs in the smartglasses space. A great deal is riding on Focals version two. Their cash position is not known, and Bosch has just commoditized their version one product. North is keeping us on the edge of our seats.

NORM GLASSES
Norm Glasses—a product of Human Capable, who began as a developer of voice-command solutions for the visually impaired—have chosen a similar product direction in a fashion eye-frames form factor, but a rather different display route, via a simple and compact prism-optics module discreet enough to hide within the eye-frame.

Weighing in at 36g, Norm Glasses are the lightest among all smart glasses with a heads-up display (even lighter than a vintage pair of classic Ray-Ban Wayfarers at 45g, which contain no electronics at all).

With Human Capable’s background in audio tech, their Norm Glasses won Best of Innovation in the Headphones & Personal Audio category for CES 2020. Their glasses are first available to demo at this year’s CES (booth 22030). While Norm Glasses are being made available now in self-branded frames, they are also seeking brand partnerships in a technology licensing business model.

GigantiCo by Chris Grayson

CHRISTOPHER GRAYSON
Christopher Grayson is a market analyst on smartglasses and the near-eye optics industry who has been published in both the tech press, and the eyewear trade press. He comes from a background in the creative department of the New York advertising industry with tours of duty at Ogilvy, GREY, and others; where he produced award winning campaigns for Intel, Nikon, and other iconic brands. In an ealier life he worked as an in-house technical illustrator at Morgan Stanley, and has since worked in event organizing including with Augmented World Expo—the world’s largest augmented reality trade event—having twice organized and curated the largest gallery exhibit of smartglasses and AR HUD display systems including AWE 2019 in Santa Clara, CA; and in 2013 at EyeBeam in New York City. Disclosure: He is currently assisting Norm Glasses as a marketing consultant.