posted by Bryon Moyer
We have already looked at some of the touch technology presented at last month’s Interactive Technology Summit. But the same touch and stylus overview presented by Ricoh’s John Barrus addressed another issue: markup.
Much of the touch technology is targeted at large format screens, including interactive whiteboards. But the key to many such devices is interactivity: multiple people in multiple places contributing to the content of the board. Doing that apparently isn’t easy using existing products: he noted that some require 40 hours of training to learn how to use. I frankly don’t know of any company that would agree to losing its team for a week so they can learn to use a whiteboard.
But such collaboration goes far beyond whiteboards. What about document review? In the old days, you printed out an original, made lots of copies that you gave to reviewers, and got back inked-up versions that you integrated. (Full disclosure… I had something of a deserved reputation for exploding red-pen syndrome when I did reviews in my “youth”… But I occasionally got as good as I gave; this scenario is dated, but feels very real.)
Now it’s more typical to use features like the editing and commenting capabilities in Microsoft Word and Adobe’s PDF readers. Those work, but if we had styluses that worked well enough to capture handwriting (which is coming), then it would frankly be easier simply to do old-school written markup electronically.
Whether real-time whiteboard collaboration or piecemeal markup, this activity may take place locally or involve the cloud, so if you took a few minutes, you could probably imagine a variety of situations where such capabilities might be useful.
The problem is, according to Dr. Barrus, there is no universal infrastructure to support this, nor are there standards that everyone agrees on. His suggestion was that we need some. The standard would have to support text and images, of course, but also strokes – including specification of width, color, transparency, end-cap (square vs. round vs. something else), and the like. There are at present four major markup languages (no, HTML doesn’t count, despite its name).
- PDF currently can handle text and images; stroke information would have to be added.
- Scalable Vector Graphics (SVG) has some stroke information, but needs such additional things as stroke timing. It also needs to support multi-page documents. It’s XML-based, so it’s easily extended.
- Microsoft has an Ink Serialized Format (ISF), but it doesn’t support text or images.
- InkML is stoke-oriented and has features well suited to handwriting recognition, but it also doesn’t support text or images.
This was largely a call to the industry for action to define the necessary standards. While formal activity has yet to commence, you can contact email@example.com if you’re interested in helping out.
posted by Bryon Moyer
At the recent Interactive Technology Summit (erstwhile Touch Gesture Motion), gesture was featured more on the day I was checking out the TSensors summit. But I did get a chance to talk both to PointGrab and eyeSight to see what has transpired over the last year.
These two companies both aim at similar spaces, gunning for supremacy in laptops, phones, and other household electronics (HVAC, white goods, etc.). Part of the game right now is design wins, and frankly, their design win reports sound very similar. So there seems to be plenty of business to go around – even to the point that it seems that in some cases, a given company is using them both. I don’t know if that’s to check them both out over time or to make them both happy or to use them as negotiation fodder against each other. To hear them tell it, business is good for everyone.
Development continues apace as well. One key change that’s happened in the last year is a move away from using gestures simply to control a mouse. Using the mouse model, for example, if you want to shut off your Windows laptop, then you gesture the mouse to go down to the Start button and do the required clicks to shut down the machine*. The new model is simply to have a “shut down” gesture – the mouse is irrelevant.
PointGrab has already released this; eyeSight has it in the wings.
I discussed the issue of universal gestures with PointGrab. There is an ongoing challenge of developing gestures that are intuitive across cultures (there aren’t many – some say one, some say two…). PointGrab doesn’t actually see this as a big issue; there’s room for everyone to acquire a simple, well-thought out gesture “lexicon” even if it means acquiring some new gestures that weren’t already used in that culture. Their bigger worry is that different companies will use different lexicons, rather than everyone settling on one set of gestures.
PointGrab has also announced what they call Hybrid Action Recognition. This is a way of making gesture recognition smarter, and it consists of three elements (not to be confused with three sequential steps):
- Watching for movement that suggests that a gesture is coming
- Looking for specific shapes, like a finger in front of the face
- Disambiguating look-alike objects
This almost feels to me a bit like yet another form of context awareness: these three tasks establish a context that says, “Hey, this is a gesture; that last thing wasn’t.” At present, this is a static system; in the future, they will be able to make it learn in real time.
Meanwhile, eyeSight noted that, in the future, you may have several devices in a given room that are gesture-enabled. Perhaps a laptop, a TV, and a thermostat. If you gesture, which one are you talking to? Well, as humans, our primary indicator is by looking at the person we’re talking to. EyeSight is looking at providing this capability as well: a device would react to a gesture only if you’re looking at it.
They’re also looking farther down the road at more holistic approaches, including gaze, face recognition, and even speech. (As humans, we can talk to someone we’re not looking at, but we use speech to alert them that they’re who we’re talking to.) But this is a ways out…
As an aside, it was noted in a presentation that gaze in particular is good for broad-level use, but doesn’t work well for fine tracking since our eyes actually flit around at high speeds (saccadic movement) – activity that our brain smooths out so that we don’t notice it. A computer could tell that we’re looking at the computer easily enough, but it would have to do a similar smoothing thing in order to be able to identify, for example, which word we’re reading on the screen.
This whole gesture space seems to be moving extraordinarily quickly; there has been significant change in only one year. This is but one reason that it’s all done in software instead of hardware; updates can be anything but minor. The other reason, of course, is that this capability is going onto mainstream consumer devices. Requiring specific hardware would introduce a much higher barrier to inclusion.
This tension between hardware and software is actually going to be playing out in related spaces, but that’s a topic for another time.
*Unless, heaven help you, you’re on the original Windows 8, in which case you’ll gesture to move the mouse all over the place in a vain attempt to find where to shut things down; then you’ll give up and gesture to bring up your favorite browser to search for “How the #@$(&* do I shut down my @(#$&(# Windows 8 machine???” and find that you go in to Settings (???????) and a few more mouse clicks (really??) done by gestures and Bingo! In only 15 minutes, you’ve managed to shut it off, with only a 50 point rise in your blood pressure! I think that, by this whole Windows 8 fiasco, Microsoft is earning itself its own specific gesture. One that I won’t repeat here, this being a family newspaper and all.
posted by Bryon Moyer
Imec has been working 2,5D IC issues with a particular focus on optimizing costs and, in particular, test yields. Yields can take what might have been straightforward-looking cost numbers and make things not so clear.
In their work on interposers, Eric Beyne took a look at three different ways of routing the signals from a wide-I/O memory. These puppies have lots of connections – like, 1200 per chip. He explored three different ways of implementing the interposer to find out which had the best cost outlook. The idea was to connect two such interfaces, with four banks of 128 I/Os each. Each channel had 6 rows of 50 microbumps. Microbump pitch along a row was 40 µm; along a column it was 50 µm. The two simply needed to talk to each other on the interposer.
The cheapest, most traditional approach is to use PCB (or PWB) technology. An aggressive version would have 20-µm pitch and 15-µm vias. This approach resulted in an 8-layer board; you can see the layout below – lots of routing all over the place. Wire lengths were, on average, 180% of the die spacing.
Next was a semi-additive copper process – more aggressive dimensions and more expensive. Line pitch was 10 µm; vias were 7 µm. the tighter routing allowed connectivity with only 4 layers, and the average wire length was 166% of the die spacing. You can see the slightly less colorful result below.
Finally, they took an expensive approach: damascene metal lines. Moving from the PCB fab to the silicon fab. But this got them down to 2-µm pitch with 1-µm vias, and that was enough to run wires straight across on 2 layers with no extra routing. In other words, wire lengths were equal to the die spacing. You can see this on the following picture.
So what happens to the overall cost? The last one is nice, but expensive to build. And here is where yield comes in. Because the “most expensive” option uses only two layers, it has the best yield. And that yield more than compensates for the expensive processing, yielding the cheapest option.
They didn’t give out specific cost numbers (they typically reserve those for their participants), but the net result is that they believe the damascene approach to be the most effective.
Images courtesy Imec.