Computers Should be, but not be Seen
With ubiquitous computing (UC), Xerox's famous PARC research center is trying to turn the way we use computers upside down. Mark Weiser, head of PARC's computer science lab, wants to make computers ubiquitous and anonymous at the same time.
Like most innovators, Mark Weiser is dissatisfied with the status quo.
"The old kind of computers, the ones that sit on your desk, require you to enter their world," he says.
"It's the computer at the center, instead of the person and their work. In extreme cases, as with "virtual reality" systems, you must put on a helmet and gloves to use them.
At PARC, which is a neighbor to Stanford University in Palo Alto, California, they are trying to do just the opposite!
The goal is an integrated computing environment, where computers are everywhere and always available. Instead of one computer on your desk, you will have dozens, of varied sizes and functions. Each room can accommodate hundreds of computers. We should not have to drag the computer with us when we go from one room to another, because we can just pick up one of the computers in the other room and continue what we were doing in the first one. Computers should know where they are and where you are and be constantly ready to serve you. This will create radically new conditions for collaboration and increased productivity in tomorrow's offices.
It is, in short, the dream of ubiquitous but never conspicuous computing.
Against this background, it may be easier to understand why Mark Weiser wrote in the September issue of Scientific American that "the very idea of a 'personal' computer" is misplaced.
"The old-style computer visionaries are trying to radicalize personal computers and make them even more personal," he said.
“Alan Kay talks about intimate computing as the next step. It makes me sick when I hear that! Getting intimate with our computers is not the right way. We want to get computers out of the way! They should be part of our life, like paper, pens, and chairs, but we don't want to get intimate with them. It's pushing the personal and individualistic to its extreme," says Mark Weiser and asks pointedly:
“Are you intimate with your pen?”
(Alan Kay was one of PARC's founders and is today one of the leading researchers in Apple Computer. He coined the term "dynabook," which Apple popularized with the term "knowledge navigator.")
Paper and pencil are also technologies, but we have stopped seeing them as such. We unthinkingly take them for granted, which for Weiser is a sign of how deeply a technology has taken root in our society.
“Ubiquitous computing is the first attempt to apply in-depth what we have learned about humans over the last 20 years," says Mark Weiser.
“It is often psychological insights that drive changes in computing. For example, screens and computer graphics are responses to input from psychologists.”
“Ubiquitous computing could only happen in a place like PARC, where you have computer scientists and anthropologists under the same roof. I came up with the concept when I was thinking about how we could respond to what the anthropologists were telling us. They studied how people use technology and talked about 'situations.' They said that we had misunderstood technology; that we had overlooked the myriad details of real situations, real spaces. About relationships between people in a room and the cultures they come from," says Mark Weiser.
This led to a radical rethink of the way we use computers. Mark Weiser and his colleagues realized that rethinking details such as the graphical user interface (GUI) would not be enough.
“Since a serious model of ubiquitous computing requires us to consider the details, we had to try to build elements of it and start using them as soon as possible. This was not something you could study theoretically, it had to be studied practically.”
“We eventually came up with some stuff that indicated where we are going. We know they may not be exactly the right things, but they are different enough from the current computers to help us along the way.“
At first glance, PARC's model of the UC doesn't look particularly remarkable. We have a large electronic drawing screen ("Liveboard"), which you can draw on with a digital "marker." We have many pen computers ("Pads") and a variety of small pocket-sized digital notebooks ("Tabs.")
The first thing that makes me raise my eyebrows a bit is a special application of the smallest piece of this puzzle: the "active" ID tag that Mark Weiser wears on his chest, which is computer with a built-in infrared transmitter. This smart badge, developed in collaboration with Olivetti researchers, is constantly communicating with small antennas on the ceiling of each room.
Smart badges can be used to both control and serve people. They can be part of a system that selectively opens doors; they can direct telephone calls and electronic mail to a person regardless of which room he is in; they can inform the computers in a room who is there.
They can also create automatic diaries, with the system continuously recording where you are (unless you put the ID badge in a pocket, which blocks its infrared signals). The system creates an automatic log of where you've been and who else has been in the same place, a list that can be a valuable memory aid. (Mark Weiser is aware that many fear that this is an invasion of privacy -- "Big Brother is watching you" – but he believes that this can be avoided through social norms about how the system is used and by allowing individuals to control such private information themselves.)
The smallest computers in the UC system, let's call them tabs, are intended to function as a very simple writing tool, an electronic scratch paper. Once could compare them to simple pocket calculators. Mark Weiser imagines a hundred such tabs in a typical office room.
The next step in the scale is "Pads," electronic notebooks. (I refrain from suggesting a translation of this term!) They are similar to pen computers, but serve a different function, as illustrated by the fact that a room can have 10-20 "Pads." Just like in Windows you can have several programs open at once, you can have several "Pads" active on your desktop. The difference is that the previous window environments are crowded on one and the same screen, while the PARC model can have them lying next to you in full scale.
Each room will also have one or two "live boards", which can be used for communication (e.g., video conferencing) or presentations.
The basic elements of "ubiquitous computing" are based on technologies that either already exist or are imminent. Lightweight handheld computers, pen computers and projection screens with a pen input system (an infrared camera records the movement of the digital "pen"), a largely wireless local computer network, and a number of networked computers.
What makes the system as a whole unique is the total integration of such a large number of computers, all communicating with each other in a uniform way and each part knowing where it is.
While Mark Weiser is drawing on his "big screen," I can give him written comments, directly from my "Pad." They will appear on his screen. I can also take the content of his screen and save it as an icon on a "Tab" to later call up the same image on another "big screen". The system is everywhere, and I can access it no matter where I am in the building. It's completely different from carrying a pen computer from room to room and connecting to a local network.
"Some people think we're going to get there anyway, more or less automatically, but I don't think so," says Mark Weiser. "Instead, we're going to have clusters of networks communicating in different niches, like pagers, mobile phones and personal computers of various kinds. Compare that to the access we have to literature in this room. We have immediate access to all the words in the room, without having to worry about what format they are printed in. As long as we (in the computer world) are dominated by these niches, we will never have "seamless" access to the world of information, where you can pick up a scrap computer and just start working.
It is important to remember that UC is still only a research project, which may change considerably before Xerox decides to try to bring it to market. First, some of the difficult technical challenges the system faces must be addressed.
"We're going to need much more wireless communication than any company today can imagine," says Mark Weiser. "We need connections between hundreds of computers in a building, whereas companies in this field think in terms of one person, one computer.
PARC researchers have therefore had to develop their own wireless networks with tiny cells, each covering a room.
Another problem is building a stable system, with so many loosely linked computers.
“You must make sure that the whole system doesn't collapse if one computer fails, something that current distributed computing systems are poor at. We need a much more robust technology for networked computers," he says.
Both computer operating systems and window management technology need to undergo major changes to enable ubiquitous computing.
The third challenge is the ability to build small. This is a prerequisite for producing sufficiently small, lightweight, and inexpensive tabs.
Even if the Xerox researchers in Palo Alto succeed in meeting these and other technological challenges, it remains to be seen how the parent company meets the market challenges. In the 1970s, Xerox managed to fumble a series of brilliant inventions, losing both its geniuses and the markets they created.
Hopefully, the 90s harvest from PARC will not suffer the same fate.
Hans Sandberg
Background:
Xerox: Not only copiers
But what does Xerox have to do with computers? Quite a lot and for three reasons.
First, because in 1970, CEO Peter McColough set up a research center to study how complex organizations use information. This center, the Palo Alto Research Center, located in a hilly area near Stanford University, would prove to be extremely creative. For example, the first Alto personal computer/workstation was invented here in 1974. Although a commercial failure, it introduced many concepts that would shape the personal computer revolution and Apple's successful Macintosh: overlapping windows, icons, mouse, bitmap graphics. PARC also led the development of local area networks with its Ethernet and was the birthplace of the first object-oriented programming language, Smalltalk.
The second reason is that Xerox for a long a time had the ambition to become a computer company. In 1987, it gave up on selling its own computer systems and exclusive operating systems in favor of Sun's SPARC computers and the SunOS operating system. Today it focuses on developing and selling applications for office automation and in particular advanced desktop publishing and image processing.
The third reason is that Xerox digital copiers are computers. They contain networks of microprocessors and advanced software, including expert technical monitoring systems.
No comments:
Post a Comment