It's not often that one gets a chance to attend a demonstration of a new method of human-computer interaction. Having been too young to witness the development of the command line in the 1950s or the modern graphical user interface at Xerox PARC in the 1970s, it was a genuine thrill to visit Microsoft's campus for a personal demo of "surface computing." While future computer historians are unlikely to view this technology as being anywhere near as groundbreaking as the CLI or GUI, the multi-touch interface nonetheless serves as an innovative way of interacting with the personal computer.
Microsoft Surface has taken many years to come to fruition. The original idea was developed in 2001 by employees at Microsoft Research and Microsoft Hardware, and it was nurtured towards reality by a team that included architectNigel Keam. Not content with merely coming up with a new idea, the Surface team is committed to actually releasing it to the commercial market as early as the end of 2007. From there, the team hopes that the product will make its way from retail and commercial establishments to the home, in much the same manner as large-screen plasma displays have migrated out of the stadium and into the living room over the past few years.
Microsoft began the Surface project back in 2001, after the idea had already been proposed by employees in the Microsoft Research division. For many years the work was hidden under a non-disclosure agreement. Keam mentioned that, although necessary, the NDA made it frustrating when Microsoft scheduled the official Surface announcement just days after Apple announced the iPhone. While both projects employ touch-sensitive screens with multi-touch capability, they are very different from each other, and the development timelines clearly show that neither was "copied" from the other. As Keam put it: "I only wish I could work that fast!"
Beyond creating the hardware, however, the Microsoft Surface team has identified several different scenarios where the device could be used in retail and commercial environments, and it has developed demonstration software that shows off the potential of the system. Microsoft has partnered with several retail and entertainment companies and will be co-developing applications customized for these environments.
Let's take a look.
Senior marketing director Mark Bolger models Surface
Essentially, Microsoft Surface is a computer embedded in a medium-sized table, with a large, flat display on top that is touch-sensitive. The software reacts to the touch of any object, including human fingers, and can track the presence and movement of many different objects at the same time. In addition to sensing touch, the Microsoft Surface unit can detect objects that are labeled with small "domino" stickers, and in the future, it will identify devices via radio-frequency identification (RFID) tags.
The demonstration unit I used was housed in an attractive glass table about three feet high, with a solid base that hides a fairly standard computer equipped with an Intel Core 2 Duo processor, an AMI BIOS, 2 GB of RAM, and Windows Vista. The team lead would not divulge which graphics card was inside, but they said that it was a moderately-powerful graphics card from either AMD/ATI or NVIDIA.
The display screen is a 4:3 rear-projected DLP display measuring30 inches diagonally. The screen resolution is a relatively modest 1024×768, but the touchdetection system had an effective resolution of 1280×960. Unlike the screen resolution, which for the time being is constant, the touchresolutionvaries according to the size of the screen used—it is designed to work at a resolution of48 dots per inch. The top layeralso works as a diffuser, making the display clearly visible at any angle.
Unlike most touch screens, Surface does not use heat or pressure sensors to indicate when someone has touched the screen.Instead,five tiny cameras take snapshots of the surface many times a second, similar to how an optical mouse works, but on a larger scale.This allows Surface to capture many simultaneous touches and makes it easier to track movement, although the disadvantage is that the system cannot (at the moment) sense pressure.
Five cameras mounted beneath the table read objects and touches on the acrylic surface above, which is flooded with near-infrared light to make such touches easier to pick out. Thecameras can read a nearly infinite number of simultaneous touches and are limited only by processing power. Right now,Surface isoptimized for 52 touches, or enough for four people to use all 10 fingers at once and still have 12 objects sitting on the table. (For more on the camera system and hardware, check out ourlaunch coverageof the system).
The unit is rugged and designed to take all kinds of abuse. Senior director of marketing Mark Bolger demonstrated this quite dramatically by slamming his hand onto the top of the screen as hard as he could—it made a loud thump, but the unit itself didn't move. The screen is also water resistant. At an earlier demonstration, a skeptical reporter tested this by pouring his drink all over the device. Microsoft has designed the unit to put up with this kind of punishment because it envisions Surface being used in environments such as restaurants where hard impacts and spills are always on the menu.
The choice of a 4:3 screen was, according to Nigel Keam, mostly a function of the availability of light engines (projectors) when the project began. Testing and user feedback have shown that the 4:3 ratio works well, and the addition of a slight amount of extraacrylic on each side leaves the table looking like it has normal dimensions.
Built-in wireless and Bluetooth round out the hardware capabilities of Surface. A Bluetooth keyboard with a built-in trackpad is available to diagnose problems with the unit, although for regular use it is not required.