304 North Cardinal St.
Dorchester Center, MA 02124
304 North Cardinal St.
Dorchester Center, MA 02124
Lucyd is offering a cryptocurrency and has to date sold about $3.7M. Lucyd is combining the hype of cryptocurrency blockchains with AR to bedazzle people. Warning, per Lucyd, citizens of the USA, China, and Singapore can’t buy the coin, probably because it runs afoul of security and exchange laws (SEC or similar).
Lucyd’s whitepaper, their website, and videos are a buzzword salad full of hype with no serious detail on what they are doing in the way of AR. They show a few”simulated” images, one of which is on the left) that could have been generated in a few minutes in Photoshop. Lucyd says (with my bold emphasis, “Lucyd Lens will seek to be the first that is lightweight, ergonomic, wire-free, comfortable, accommodates corrective lenses and appears like a normal pair of glasses.” Yeah, everyone seeks this, but then these other companies have a lot of serious technical people working full time on the effort and not just bunch of “advisors” as listed in the white paper. It looks to me that they licensed a bunch of old worthless patents to give them credibility. I will be going through the patents later; they are hilarious).
I have seen others raise questions about the soundness of Lucyd’s cryptocurrency (search “Lucyd scam,” but I have not investigated this issue. This article is meant to expose the patents at the rotten core. Other online magazines (including Forbes) just take for granted there is some “there-there” regarding technology, I bothered to check.
As a teaser, there is a connection to Magic Leap toward the end.
Quoting the Lucyd white paper:
Lucyd is launching the LCD token generation event to decentralize the AR revolution. We are designing next-gen smartglasses that we believe correct many of the issues plaguing currently available products, and a blockchain app ecosystem to support them. Lucyd Lens smartglasses will be designed to evolve your vision, by seamlessly merging valuable data with your sight.
The Lucyd Team, Headquartered in Singapore (although none of the principles seem to live there) is heavy on MBA and Marketing and very light on technical people and does not even list a CTO. The chief optics person, Dr Mike Kayat, is also an MBA, and it looks like he started in sales and marketing and then moved into management with NO serious technical work (no patents and no published papers I could find). There are ZERO people listed other than as “Advisors” that have any background in optics or display (why worry about technology when you are selling “coins”?).
CEO Dr. Gross profile shows he is also CEO of Tekcapital, of which Lucyd is being spun out. Before that, he was the CEO of UTEK another patent licensing entity. His published background is a CEO of “patent licensing companies Tekcapital and Utek. He has a few random patents in human factors but nothing to do with head-mounted displays or optics. This makes me wonder if the ICO is really to fund some form of non-practicing-entity (NPE) to sue people over what appears to me to worthless patents.
The white paper includes a buzzword-laden list of “Anticipated AR Features.” If anticipated was not enough of a weasel word, they added for good measure, “Please note: actual technical features may vary.” It looks like someone just looked through the patents Lucyd licensed and randomly pull words from the different and incompatible patents. I’m going to try and break some of these down.
Take for example under light guides, “1. See-through phase conjugate retroreflective material (semi-transparent curved mirror).” Are they doing retro-reflectors or a simple curved mirror? It appears that an incompetent person was doing the cutting and pasting with no understanding of what the terms meant. The term “phase conjugate retro-reflective material” probably sounded high tech to the technically illiterate person doing the cut and paste to make this table. Phase conjugate material is a retro-reflective screen and thus not semi-transparent. The patents even call out “phase conjugate material screen 504 (purchased from 3M Corporation),” which is that highly reflective beaded material 3M makes used in things like signage and reflective wear., but what it is referring to in patents 6,804,066, 6,963,454, and 6,999,239. They pulled the “semi-transparent curved mirror” from different patents.
Number 2 under Light Guide is random gibberish made up of words from several different patents thrown together. The got “the cube” part from X-Cubes (patents 7,119,965, 7,639,208 7,843,642) or beam splitting cube (7,522,344), none of which are using TIR (total internal reflection). The micron-level part were microlenses from patent 7,009,773. And the TIR came from a patent 7,843,642 about fiber optics “Where a bare optical fiber is used in a total-internal-reflection (TIR) capacity.”
Under “Microdisplays” they mention OLED (no idea what the ~2′ implement means). Ok, the resolution of a 1280×1025 is a typo, but the resolution of 1392 by 1040 is not any display that I know of and was hilariously is copied from table 2 (left) of 7,522,344 from the camera (not display) resolution. Obviously, somebody was just cutting and pasting numbers they did not understand.
Under projection optics, they say “1. Compact, lightweight telecentric lens designed with free form optics (aspherical and DOE).” They have combined words from various patents into gibberish. “2. Tiling method to achieve very large fields of view (FOV) and high image display resolution” is from patent US 7,119,965 and is silly (as I will show later).
They claim to have “1. Demonstrated designs with 1024×1024 & 1392×1040 microdisplays.” which appears to be a lie. I know of no company that has made a 1392×1040 resolution microdisplay, but this is a camera sensor resolution (see table 2 above left). The claim appears to be made by someone cutting and pasting information they have not even read.
From what I can see in the 13 patent that Lucyd has a licensed are worthless other than being used to deceive potential investors. I am going to show what I consider the key figures from all 13 patents below.
Both of these related patents have a specific projector lens design (Fig. 2) in a bizarre configuration (Fig 1) for projecting into a room with a retroreflective screen (“phase conjugate”). See Figure 1 of patent 7,119,965 Head Mounted Projection Display with a Wide Field of View below for an idea of how this is to be used (it is hilarious).
This patent is on a very generic method to detect if a person is drowsy (see flow chart on the left). There are a lot of things you have to check be able to check for, and I seriously doubt they are the first to use a camera to detect a person is drowsy.
These three patents are all related to the same original filing with just slight variation. They use a bizarre optical configuration with some “phase-conjugate” material (retro-reflective 3M sheet) that nobody is ever going to use in a product. Note the phase conjugate material is at the bottom under either a Fresnel or microlens (this are the “micron size” thing, but they are not “corner cube arrays” per the “Features”). Hey, but the headset is stylish (heavy sarcasm)?
This patent is an algorithm for detecting pupil shape. It is one of many a nearly infinite number of ways to that could be used and thus easily avoided if one were doing this.
The keyword in the title is “projection.” This is a projector that you mount on your head and then have a retro-reflective screen (more of that “phase conjugate” material) as illustrated in Fig. 1 (right).
They have a very complex optics cube for allowing two microdisplays to be combined to support projecting a wide FOV on the retroreflective screens shown in the side and top views.
I don’t think anyone building a product is in any danger of using this patent. This is Lucyd’s tiling patent for wide FOV, and I think I can say without fear of rational contradiction, the concept is absolutely worthless.
Below are Figures 2, 4, 5, and 9 (2 and 5 are inset). The optical configuration of Figures 2 and 4 won’t work as they draw the light rays incorrectly off the semi-transparent curved mirrors 304. I have added with dotted lines how the rays will miss the eye. Interestingly this related to my last article about the Intel Laser Beam steering where they used a hologram to bend the light at a sharper angle to direct it into the eye.
They seem to try and fix the error in Figs 2 and 4 and fix it in Figs 5 and 9 by adding a partial (semitransparent) curve mirror 506/516 at a steeper angle along with some corrective optics 504. But this configuration will lead to terrible distortion of the real world as you are left looking through a curved lens at a severe angle. In short, both designs are useless for making a decent headset.
With this patent, we yet again have a system that depends on having retro-reflective material. In this case, it is part of an eye tracking system. I suspect that this is to go with the some of the other systems that have retro-reflective screens. This system mixes IR for eye tracking (something people have been doing for a long time) with using the retroreflective (or if you prefer, phase conjugate) material.
This one is mildly interesting from an R&D perspective but among other problems requires it requires a component that does not exist (as far as I am aware), an X-Cube Beam Spitting Polarizer. There are X-Cubes using dichroic (color) filter and there are beam splitting prisms, but there are no X-Cube polarizers (I’m not sure they would even work as they intended). The patent goes on to tell what they actually built on the right that only demonstrated occlusion but not the combination of occlusion with a virtual image. This is a case of “insert magic component here and my idea will work.”
It so happens that ASU in 2017 published a paper related to patents granted in 2017 by their researchers and license to Magic Leap on occlusion that shows what it actually takes using components that exist. Of further interest is that Hong Hua, who in the paper states she is a consultant for Magic Leap, is also named as an inventor on Lucyd license patents 6,731,434
and 6,804,066 [correction] above and has papers mentioned in others.
I should add that I would not be that worried about the ASU occlusion patent licensed to Magic Leap. It will take a while to explain but it only works for the special case where the “real world” is nearly flat. “Hard-edge occlusion” for the general case with depth in the real world is one of the “Holy Grails” of AR and may not be possible (hope to cover in another article).
What can I say? This is hardly compact and nobody I know of building a real headset would do this. Specifically, I don’t see people using a “hybrid reflective structure (a fancy name for a mirror light pipe) going to a small X-cube. This would be ridiculously expensive and large for no added benefit.
These patents appear to be is a bunch of graduate student ideas that are impractical and that nobody in the industry is using in the last 8 to 17 years since these were filed. I believe their only value is for a dishonest person to throw them in a convoluted cryptocurrency scheme to deceive people into believing that they have some serious intellectual property. Unfortunately, the AR space is so hyped up today and magazines and online publications just repeat what they are told.