Tags:

We recently had a chance to sit down and talk with the two of the founders of emerging startup, ThinkOptics – Gopal Panchanathan (CEO) and Anders Grunnet-Jepsen (CTO). San Jose-based ThinkOptics’ approach to motion control differs from most competitors in that it is a wholly optical solution, without the inertial (gyro, accelerometer, etc.) tracking that is all or part of competing solutions. We had a chance to test out their system at CES where they were exhibiting that product for the first time, and recently followed up with them post-CES.

Motion Applications Report: How big is ThinkOptics, and how are you funded?

ThinkOptics: We started the company two and a half years ago. Three of us started the company, ourselves and John Sweetser, VP of Engineering. Beyond that core, we’ve got a group of consulting engineers with whom we contract for development, and an Advisory Board.

We launched as a self-funded company, and have since taken on an angel investor. We’re currently in the process of seeking Series A financing.

Motion Applications Report: What market you are pursuing?

ThinkOptics: Pointing devices, navigation of media, digital entertainment – any market in which a solution vendor brings a set top box or similar device to bear, or as part of the whole foodchain… So our focus is the so-called ‘ten feet interface’ between the TV and the sofa… it could be part of a solution from any number of players in this space - Apple, Microsoft, AT&T, Comcast, or a gaming subscription solution. We’re focused on anyone bringing convergence to this market. We are not tied to Media Center Edition or any specific UI. We can integrate to any platform – we are offering it on Windows, Mac and Linux.

Motion Applications Report: So does this include the remote control market?

ThinkOptics: Yes, with a huge difference – we believe one should not have to look down at the remote at all, in order to control what’s on screen! The current remote control paradigm is crumbling. We are just astounded by those companies who are competing on how many buttons to put on the remote. But there’s a backlash going on. Not too long ago, I was at a conference where a guy from a well-known set top box vendor boasted that his remote control had come down from 60 to 30 buttons. He was waxing eloquent about how they had solved the 60-button problem. But what about the 30-button problem?!

When you have a large screen, the last bastion of complexity to fall is the user interface and remote control. Ease of use – that’s what is going to set the winners apart from losers in the living room. We see urgency today in vendors trying to solve this. On your large flat screen TV, it should be just as easy to surf with Opera, the web browser, as it is to watch Oprah, the a TV show.

Motion Applications Report: How big do you see that market being?

ThinkOptics: How many people are going to want to sit on their sofa and enjoy a way of intuitively accessing their TV? That’s our market. Nintendo Wii with its direct pointing user interface access has helped seed this market and has made consumers aware of the benefits of “direct pointing” way of access. We bring this sort of intuitive human access to the TV near you - and every one of the 110 million households in the US is our target customer in the US.

Motion Applications Report: What are your channels to that market?

ThinkOptics: We’re pursuing set top box vendors, Consumer Electronics vendors, and IPTV vendors. We’re also working with the other people in the value chain, for example middleware vendors for IPTV. Another potential channel for us are the integrators in the IPTV space, who bring a total IPTV solution to a carrier.

Motion Applications Report: Tell us how your technology works?

ThinkOptics: It’s optical pointing. Our system consists of two things – an optical sensor in the handheld and, near the TV, what we call ‘the Pod’. The Pod has the “markers”, the handheld has the sensors.

We have no MEMS or accelerometers in our solution, it’s all optical-based. With this solution, we get an extremely high pointing resolution (a capability we call “Direct-Point”), the ability to discern with great accuracy the angle of twist of the wrist (“virtual-dial”), and, “easy-zoom”, a capability in which the system can tell precisely if you are moving the wrist forwards or backwards. We can also detect where you are sitting in the room, which is real useful to know in some cases. Finally, we also support multi-player capability, where up to four users can simultaneously be recognized as unique players.

There are several ways these four capabilities can be used within an interface. First is the direct-point capability. You point at the screen, and where you point is where the pointer goes. It’s like a laser pointer without lasers. If you move beyond the TV screen boundary, we can detect that too. You can use this “off the screen” for scrolling – say, scroll up by just pointing below the TV!. Direct pointing is the most important part of our solution. Pointing is the most natural way we humans evolved to use our hand and wrist in interacting with the natural world. One more thing about direct-pointing: unlike laser-pointing, we can correct for the natural shaking in the human hand, through our built-in stabilization, much like in a camera.

Second, we support a “virtual-dial”, the twisting operation when the wrist can twist right or left a few centimeters, and make things happen intuitively. This can be used, for example, to control volume like a dial, or move forwards or backwards while browsing the Internet. You can also use the twist to “jog” forward or back while watching a movie and for pause and play functions as well. All six jogging buttons can be mapped into just one simple twist of the human wrist!. Only your imagination is the limit in how you could envision using the simple twist to make TV watching a simple experience.

The third capability of the system is what we call “easy-zoom”. This detects how far you are away from the screen. By moving your wrist forwards and backwards you are able to zoom in and out. For example, if you are looking at pictures it can detect your wrist motion forward and back to zoom in and out – say, you want to zoom in on grandma’s face in a family picture. So this brings a third dimension to the interface.

The fourth aspect of our solution is that it is inherently multiplayer – we are able to support many simultaneous users accessing the screen. This is very important for multiplayer games. The system knows each player distinctly from the others, something that’s not possible with inertial devices.

Users can pull up keyboard on screen, … a virtual keyboard. They can also feel icons on the screen with haptic feedback.

Motion Applications Report: How does ThinkOptics compare to other approaches being taken in the market?

ThinkOptics: We’re familiar with Gyration, etc. We admire how they are in the forefront of trying to change people’s perceptions. But we advocate and believe in direct pointing. Our users don’t have to worry about where the cursor is right now … they don’t have to worry about finding the curser and then moving it. Our device cannot get ‘lost’ on the screen ─ when you pick it up and point it, you know where it is. That’s not the case with inertial systems, where, for example, if you tilt your hand 90 degrees and you go left and right – the cursor moves up and down – because it has its own frame of reference.

Fitt’s Law is a well established way of measuring how good a pointing device is. In our testing, we equal or, more typically, beat the mouse. No inertial device we have tested – and we have tested a lot of them! – even comes close!

Secondly, the inertial technologies need to add new devices (such as additional accelerometers) for new axes or degrees of freedom. So to be able to add the ability to rotate, tip and tilt, you need to add new sensors. In our system, it’s one shot – you get everything you need, from the same basic configuration.

So to be more specific with an example, with gaming, you want to point and shoot. The inertial devices all have a relative frame of reference. While other devices measure a change in position, we measure actual position. We can differentiate once for motion – differentiate again for acceleration. With inertial motion sensing you can’t integrate back to learn about the exact position.

Additionally, drift is an issue for these guys [the inertial motion sensing players].

With sub-millimeter accuracy, we can detect how close to the screen you are. With a very small motion or action, we can detect what you are doing.

Motion Applications Report: Who are your main competitors, as you see them?

ThinkOptics: Anyone who’s got a pointing device in this space – anyone who offers easy access to information. There are only two approaches – the inertial approach or optical approach. We are the standard bearer in the optical approach (not including Wii and its “sensor bar” – which appears also to be optically based). If you love the Nintendo Wii’s direct-pointing approach, and you’ve got a Linux platform, we are the only game in town.

Gyration, Hillcrest (both of which we think are very good) are on the OTHER side of the technology–namely, the inertial approach. We think this has disadvantages in rapid access, which is crucial for gaming. Most of all, this approach does not support direct pointing as we do.

The optical approach is extremely accurate compared to inertial. Speed is computable via Fitt’s Law, and we find we are faster, through this objective, quantitative measure.

This is the Holy Grail for console gamers… as the mouse and the keyboard are the Holy Grail for PC gamers. Inertial devices don’t come close to optical systems in terms of Fitt’s Law characteristics. Our device even exceeds the mouse and keyboard.

We’ve measured Gyration and mouse and keyboard … we’re at par and better than mouse, and much better than Gyration. You’ll find that none of those guys are in gaming, and there is a good reason for it.

Motion Applications Report: How is your solution integrated into devices? Into software?

ThinkOptics: We can be integrated into a traditional remote, as well as a PDA or cell phone. The Pod that sits near the screen, can be integrated into STB, console, IPTV, or it can be a standalone device.

In terms of software integration, we have an SDK, which can help map the UI functions onto the device.

We’re beginning to do STB integration, it’s not done yet. Processing power required for our solution on a set top box is 1/3 of that required for a mouse. On a PC or Mac it is running on “0%” CPU, as reported on the Windows Task manager.

We can easily integrate into any middleware. For our existing integrations which we show in our demo, we haven’t had to change the application UI say, in MCE or Opera, in any way – we just map into those platforms and transform the user experience into something much more “human”.

However, the best approach/end result will be when people use our interface “natively”, that is design the UI around the capabilities we have built into our device.

Motion Applications Report: What is your cost basis compared to inertial solutions?

ThinkOptics: We’re comparable, but we believe we are a lot cheaper when you consider adding additional degrees of freedom [to the inertial systems]. Everything comes standard with our solution.

Motion Applications Report: Where do you expect to see the fastest adoption of your technology?

ThinkOptics: All of the markets we’re chasing…,there’s low hanging fruit in all of them. We can instantly go into a MCE and transform it. Access and navigation of MCE through our device is 3 to 4 times faster than traditional remote control that ships with an MCE, based on our informal tests. This is because we already know how to intuitively use our wrist. One of the most underutilized device is the human wrist, always available near the remote control, and we simply “recruit” it for use in our system!

PC vendors and remote control vendors are the people who we are talking to most today, and this is changing fast, as IPTV and set top boxes that serve convergent content are becoming more common.

Motion Applications Report: What’s the role of standards in this? Can you see standards for motion applications? If so, what would they be?

ThinkOptics: Don’t really see any standards issue. We use ZigBee for connectivity between the remote and the device, but can easily switch to other protocols as needed.

There are no standards in UI interface because every one thinks they have the best. Our SDK can integrate with anyone else’s UI today. We are agnostic when it comes to UI.

Motion Applications Report: What are the biggest hurdles to your being successful, from an industry perspective?

ThinkOptics: The biggest industry hurdle is the quick adoption of direct pointing as the way to go. Until Nintendo came along, there were a lot of questions about direct pointing.

Wii had changed the game. What would have been a big barrier has now been removed. We can now bring the Wii capabilities to an open system.

Motion Applications Report: Who’s the best strategic partner you wish you had?

ThinkOptics: If there are strategic thinkers from Sony who want to distinguish themselves, and win in the video game battle, they ought to be talking to us. We are the only game in town if a Sony or a Microsoft, to name two names at random, have any intention of competing in this increasingly Wii-dominated world of gaming for the non-gamers. Our optical approach, direct pointing and all the rest of it, can truly help them bridge the gap that exists in the play experience, or even launch a product line for the non-gamers.

Motion Applications Report: What’s the main message you’re trying to get to the market?

ThinkOptics: The message is simple: If you want to have the kind of ease of access to the UI like Nintendo Wii has, you need direct pointing. And if you want to reduce stress on the user, and improve the ease of use (the elephant in the room for many systems), you’ve got to talk to us.

You can try as much as you want with inertial devices, they’re much harder to use for all the reasons we have discussed. What you need is not just a mouse on the couch. You need something that can do what a mouse can do, PLUS other things. Something that supports all of the freedoms that the human wrist provides you. You need go beyond the two dimensionality of a 3D-Mice/3D-remote control.

Our main takeaway is this: The ten foot gap has to be bridged by human intuition. Nintendo Wii has shown that direct pointing is the way to go. Human wrist is there for us to use, but much neglected by the traditional remotes. Why not take advantage of this wondrous human-device, produced by eons of evolutionary design? We provide that and more, to all open platforms.