Author’s note: Portions of the research for this project were provided by reader K. Thank you for your help.
Every few weeks, we see another crowd funded campaign trying to revolutionize the way we look at our wrists to see the time. While most of these products aim to add new features to a standard watch, Ritot claims that they can display the time in a whole new way. Not content to look at the simple display of virtually every watch ever made, Ritot is trying to take it a step further with a watch that projects the time on the back of the hand.
Presumably, using the back of a hand provides a larger “screen” for displaying the time and other information, but anyone who’s tried to read a PowerPoint slide in a sunny conference room can think of at least a dozen problems with this solution. Instead of providing demonstrations of their unique solution, all Ritot has managed to produce is misleading graphics and shady details about their level of progress. This is one campaign where I wish I had a chance to dive in earlier, but with just three days left in the campaign, it’s better late than never.
In several updates, the team has informed that they’re planning on using a Digital Light Processing (DLP) projection solution from Texas Instruments. DLP involves an array of thousands (or millions) of microscopic mirrors vibrating back and forth. These mirrors each represent a pixel and will either reflect light towards the desired surface, or reflect it away. By varying the color of the light in time with the vibrations, a colored image can be created.
There really isn’t too much to say about this. The team goes into detail about why they chose DLP in particular, but when it comes to projecting images from small projectors, there are a number of tried and true options providing various levels of brightness and resolution.
They do seem to have put some thought into their design. One of the concerns with any kind of projection anything is being able to get enough contrast. A projector cannot display “black”, it can only hope to display a white that is bright enough to make the ambient light hitting the screen look black by comparison. This is problematic in the aforementioned sunny conference room, but it’s a whole different problem when it comes to reading a watch in broad daylight.
The intensity of the light incident on a surface is known as its “Illuminance” which is often measured in units of Lux or “lumens per square meter”. According to Wikipedia, Sunlight can provide between 32,000 and 100,000 lux. Seems like a tall order to compete with that. The team claims that they will be using a 10-20 lumen light. That’s not a huge number, but remember that a lux is a lumen per square meter. If the projector is only meant to shine on a 30x40mm display, 20 lumens of light derives over 16,000 lux.
Though this is larger, it’s still not going to be bright enough avoid being utterly washed out in direct sunlight. Even dropping out of direct sunlight, the amount of contrast available on a sunny day will make the watch difficult to read at best. One white paper I found on the subject of measuring and calibrating front-projection televisions claims that a PowerPoint can be viewed with a contrast ratio as low as 5:1. Assuming seeing the time is akin to reading text on a slide, that means that the watch will not be visible until ambient levels drop to 3,200 lux or somewhere close to an overcast day.
That isn’t bad! Most projector manufacturers are trying to leverage projector technology to create large images on a wall, but by utilizing a projector in a small screen space (where a conventional LCD arguably makes more sense), this kind of design might actually be visible in many lighting environments.
Now that also depends on the reflectance of the “screen” material. Someone with very dark skin may find it especially difficult to see. It’s no wonder they chose primarily white and asian
models stock photos to show off the renderings of their product.
Now, that’s probably the only positive thing I have to say about this thing. Indeed, when used over very short distances, small projectors can produce legible text. Now let’s talk about some issues.
In order to project an image from a watch to the back of a hand, the light will need to be incident at an extremely narrow angle. If the watch is indeed 12mm thick, then the light hitting the far side of the display (let’s say it’s 100mm away) will be coming in at a 6.8 degree angle.
That’s a very sharp angle.
This is problematic for a number of reasons.
Firstly, at an angle that narrow, a large portion of the reflected light will continue in that direction away from the user’s eyes. You can think of it as a stone skipping off a lake. The contrast calculation above assumes that the viewer and projector are facing the surface at relatively perpendicular angles. With the projector at such a narrow angle, the visibility is sure to be atrocious. This is particularly bad if the hand is sweaty, greasy, or otherwise shiny.
Here’s an image of an LED illuminating the back of my hand from 12mm above my wrist:
As you can see, the back of my hand, like many hands, isn’t flat. With the light coming in at such a sharp angle, even small ridges are bound to cast shadows and distort the text projected by this watch.
On top of that, using such a device requires your hand to be at a very specific angle. Bending your wrist down by just 7 degrees will cause some portion of the image to never hit your hand, and angling it up will cause distortion. Further adding to this issue is the fact that the bangle-like shape of this device doesn’t offer any features to prevent the device from rotating on the users wrist and missing its target entirely.
We’ve already established that the projected image is brighter as the screen gets closer to the projector, but what about when portions of the screen are at different distances from the projector? The amount of light incident on a surface should decrease with the square of the distance from the projector. A target twice as far away will get 1/4 as much light per unit area. So if this thing is projecting on a hand, the portion of the image closest to the wrist will be many times brighter than the portion at the end. This may be fixed in software by adjusting the image to dim the portion closer to the wrist, but that kind of dimming will require some pretty specific tuning that likely won’t carry over from user to user. Furthermore, bending the wrist as mentioned above will alter the distances significantly and present a poor image.
Here’s a big one. Projecting an image from an extremely short angle is not an easy task. The amount of specialized lensing required to create an undistorted image is very bulky and expensive to develop and produce. Projectors that shoot at extreme angles are called “Ultra-Short-Throw” and have many applications in providing large displays in cramped spaces where the display can be easily hidden when the projector is shut off.
At the moment, the champion of ultra-short-throw is Sony who recently unveiled a projector which can produce a 147 inch 4K image on any white wall:
While it’s pretty, it comes at a price of somewhere between $30,000 and $40,000. The real take-home here is that even this projector with all of the force of Sony behind it still needs about 20 inches of space between its lens and the wall, and it can only produce an image that’s around 80 inches tall. That works out to a 14 degree angle or over twice the angle of attack of Ritot.
So how is Ritot planning on doing it?
Let’s hope they remember to show us some time in the next three days.
The team has been pretty vague on their product’s features (more on that later), but they have been very specific on its dimensions:
That’s one sexy bracelet.
Elsewhere in their campaign, they provided some details about what’s “under the hood” and offered up this image:
It’s certainly an impressive array of components to fit in such a small space. Actually wait, does it fit?
The DLPC300 is a chip offered by Texas Instruments that makes it easier to interface with their DLP pico projectors. Dealing with video signals typically involves a lot of electrical connections, and this part has 176 such connections that limit how small it can physically be. How big is it? The data sheet provides a diagram:
So it’s between 6.9 and 7.1mm on a side. If the whole band is only 22mm across, it sure does look awfully small in that picture. Let’s use that chip to set the scale and do some measurements here:
Wow! It looks like the band is 45mm wide! That’s over twice what they’re advertising and would turn this thing into some kind of crazy wearable tube. Every component on the board has the correct scale relative to the other components, but it looks like they just shrunk everything down by a factor of two to fit it into their design. It’s possible that this is just a quick and dirty CAD drawing just to give an idea of where the circuit board will be going (after all, it looks like it was done in Google SketchUp), but they’ve used this scale in a number of places such as their about us page:
You could call this deceptive marketing, but the average backer probably isn’t going to even know or care what they’re looking at, and anyone who does know will immediately spot the error. Maybe if they rearranged a few things, they actually could fit the circuit into the band. But then, what about the battery?
Typical lithium polymer battery packs have a maximum energy density of 300 watt-hours per liter. 1500mAh in a 3.7V cell is 5.55 watt-hours requiring a minimum volume of 18.5mL and that’s not counting the foil sealing required around the battery packs or the reduced energy density of curved batteries. Their wristband is 12mm thick and 22mm wide. Allowing for 1mm of housing material around the battery gives you a cross sectional area of 200mm^2. In order to fit a 1500mAh battery into those dimensions, it would need to be 9.25cm or 3.6 inches long. Assuming the absolute best conditions, roughly half the bracelet will be made of battery leaving very little room for the projector and associated electronics. Even if you got that to work, what about the “sport” model?
Where does the battery go here?
The point is that while these guys might have done a little bit of research regarding component selection, etc, they certainly haven’t done even a modicum of feasibility work. The 3D model of their circuit board was not generated by any kind of circuit design package, and was likely just hobbled together by whoever did the rest of their 3D modeling work. This level of laziness is widespread throughout their campaign.
Lack of Focus and General Laziness
So okay. It’s a watch. They call it a time piece. It even comes with a charging stand that allows you to change the time and color as demonstrated in this roller coaster of a gif:
Alright. Now it also can do notifications from your phone:
How do you set up those notifications?
So why do you need a base station to set the time? Adding a full color curved touch-sensitive display to a charging station seems like an awful lot of effort for what could be done easily through the app. Also there’s this:
More than 20 colors? If they’re installing a DLP projector into this thing, it should have thousands of colors. Why stop at 20?
What does this app look like?
How are they going to waterproof this watch?
Are they making up features as they go along?
How come they’re using a battery manager designed for solar cells to charge a battery in a wireless power application? Where is the wireless power antenna? As I’ve explained elsewhere, metal objects more or less cancel out wireless charging transmitters. Ritot has an metal band that butts up right against the charger.
Moving on… they’re using an analog accelerometer for things like tap or shake detection. That means that their main processor is going to have to be listening to a live feed of data from the accelerometer. Furthermore, the accelerometer will need to be awake all the time drawing 2.1mA along with the entire application processor, which draws something like several hundred milliamps. This feature alone will kill the battery in a matter of hours even if you never turn on the projector.
Why not use something like the LIS2DH? It’s not very expensive. It has an integrated tap-detection processor that draws at its worst, just 185 microamps or roughly 9% of what the Ritot accelerometer draws, and the application processor doesn’t need to be turned on while it’s running. Oh yeah, and at just 2x2mm, it’s lot smaller.
Like seriously. A lot.
How much thought could they have possibly put into this component selection?
I know everyone loves to jump on the “scam” bandwagon, but I really don’t think these Ritot guys are confidence men. If they were, they wouldn’t have bothered providing so much detail. Their work so far indicates a small amount of research has been done, and they do certainly seem eager to show progress with some of their more recent updates.
Unfortunately, based on what I’ve seen, I have to conclude that these guys have no idea what they are doing. I speculate that this may very well be their first ever electronic hardware project or near to it, and it is an incredibly complicated project even for a highly skilled team. So far, the only real progress they’ve shown is two 3D printed plastic bracelets:
And a “demonstration” using a Texas Instruments development kit that they just received just three weeks ago:
If you want to do this at home, you can purchase the projector here for $600, install the associated software (as demonstrated on their laptop screen above), and follow the user guide. It’s not rocket science.
This is really just a clear cut example of a few guys with a neat idea grossly underestimating what it takes to develop a product. With their level of experience, I have to expect that none of the 6,492 backers will ever see the watch that they paid for.
And you know what? You can get angry, but it’s not really their fault. We’ve seen extremely inexperienced people take on projects like this before. Who can forget the infamous Project Aire which shut down with these humble words:
after having raised just a few thousand dollars. Even if they did complete their campaign, Aire would have only lost a handful of of strangers’ money.
The Ritot team has no prototype, no skills, and no clue what they are getting themselves into, but that should be okay!
That’s what crowdfunding is all about. The premise of this campaign is legitimate; these guys want to try to deliver a very complicated project, and they’ve made every effort to show progress and keep communication channels open. They’ve even openly admitted that they are in the very early stages of development:
The fact that they have over a million dollars in funding and will soon have a gaggle of angry backers is just a result of mismanaged expectations. Of course someone asking for money is going to give you every assurance that they can pull it off. They have to. The problem is the overwhelmingly positive attitude in the crowdfunding space. If there weren’t a huge pile of headlines like this:
These guys could be burning just $10,000 and slowly fading into obscurity instead of facing the horror of thousands of pissed off backers.
Hopefully the eventual fallout from this project and others like it will cause those infamous words “Kickstarter is not a store” to sink in. Skepticism is important, but at a bare minimum just a moderately level headed assessment of where money is headed would prevent tragedies like this from unfolding. It’s a shame that at least $1M will have to disappear before the general attitude changes, but I’m certain it will.
I’d just hate to be this guy in six months: