Ben Kudria » Weblog

Today, in December 2007, of the 6.6 billion people living on our planet, around three billion have access to cellular service (1).  In contrast to the relatively staid growth of fixed telephone lines (fig. 1), the growth in cellphone and Internet service subscribers is staggering.  This growth is evident not only in the United States and Europe, it can also be seen in second- and third-world countries such as China, India, those in Africa (2), and those in the Arab world.  Although these countries are not as wealthy as the US, and certainly not the center of technological developments and innovation, their governments are finding that cellular and Internet service are basic and vital infrastructure that their populations highly demand.  Almost half the population on Earth is connected somehow to a telecommunications network, either the Internet or the local cellular network.  Furthermore, this proportion is on a strict upward trend.­

 

While he was serving as Chief Technologist at Xerox PARC, Mark Weiser coined the term "Ubiquitous Computing" (3) to describe the pervasive presence of personal computing devices, connected by always-on networks, and serving humans in the most natural and "invisible" way.  The modern use of cellphones, and in a much lesser sense, personal computers and servers, is the start of the realization of Ubiquitous Computing.

While the incredible rise of cellular connectivity in Africa is staggering, it only serves as the precursor to a much more important technological progression: the broadband Internet connection.  A reliable and fast connection to the global network we call the Internet is an essential aspect in Ubiquitous Computing.

Ubiquitous Computing requires that each person have at least one personal computer that they use to supplement their everyday life.  Whether that person is a Moroccan hunter coordinating the hunt for food with his cellphone, or the savvy Manhattanite using their cellphone to find the closest sushi spot with the highest reviews, they are using their cellphone transparently to live their daily life.  As technological growth continues, these cell phones will become mobile personal computers.  They will be smaller, faster, and smarter than cell phones, and they will be used by everyone to talk to everyone.

Vital to this dream is a fast, wireless, and always-on connection to the Internet.  Local networks will exist for a while where it is difficult to get a global connection (i.e., children in a rural village using their XO laptops, or cars on a fast-moving highway connecting to provide traffic updates) but eventually these local networks will merge with the Internet permanently.  Would you rather ask the local block about the best sushi place, or anyone who has ever eaten in Manhattan?

It seems that the ubiquity of a personal connection to the Internet is rising much faster than the advancements in physical hardware, which I think will give rise to an interesting trend.  In the near future, these personal devices will not become small enough quickly enough to keep up with demand.  As an intermediary step, an industry will spring up selling "personal servers", probably virtualized.  The purpose of such a personal server will be to offload the computing power from the personal device to the "cloud".  They will serve as personal email servers, and as hubs to your personal social network web service.  Your local personal device will establish a connection to your trusted "gateway" server first, for greater reliability and security, and from there, perform any tasks you require.  As hardware size and costs shrink, however, these "servers" will shrink and eventually move to your pocket - there will not need to be a separation between the access device and the computing service.

Ubiquitous computing will greatly enhance amount and quality of reference materials available to everyone, and this will have very interesting social implications.  Already we are seeing the importance of a community-edited reference in Wikipedia, as well more niche, and more subjective references, such as Yelp.com for restaurant reviews.  Ubiquitous computing will trigger massive growth in these services because people would be able to access them more easily, and more significantly, update them more easily.

We will also see the rise in information aggregators, which will collect all the information about a topic in near-real time.  For example, on October 30th 2007, a pretty large earthquake occurred near San Jose in California.  The news networks would take hours or days to inform anyone who was tuning in, but many people heard about the earthquake seconds or minutes after it occurred, directly from friends or family, through a service called Twitter (4).  The same phenomenon occurred with recent forest fires in southern California, and anyone could view updates about the fires by searching for the tag sandiegofire (5).  In the future, we will see a single aggregator source that will show personal updates and comments, news coverage, recent photos by amateurs and newspaper photographers, and anything else related to a particular event, phrase, "tag", date, or person.  Such a site would be the first stop for breaking news.

However, ubiquitous computing is not just about querying the network anonymously.  Ubiquitous computing is just as much about the individual as it is about the wisdom of crowds.  At first, they will simply serve as a more advanced address book/social network node and calendar.  After that, I think, these personal devices will also serve a much more personal and individual need.  They will record every conversation you have with everybody, and make note of whom you spoke to, what times, where (GPS will be an integral part, obviously), as well analyze, index, and link the conversation to every other conversation you have had in the past.  You will be able to search and view all this saved material, as well see all the conversations you have had with a given person, everyone you talked to about a given topic, and what your conversations were about, on average, three months ago.  Research is already being conducted in this area.  Gordon Bell built such a ubiquitous monitoring device (although a very bulky and primitive variant) to record audio and video of his entire life.  Dumping this data to a computer, and then combining it with everything he has done on the computer (i.e., all the email he has sent, every website he has visited, every IM conversation, etc), the software is able to make useful and sometimes surprising predictions. (6).  Do you ever want to tell someone about something captivating you noticed, or heard somewhere.  With such deep, complete, and systematic access to a database of all information you have encountered, you will be able to bring up relevant text, video, and other primary sources within the second.

However, there are some problems we must surpass to make ubiquitous computing possible.  These range from problems in power supply and battery technology, computer interfaces, software and software engineering hurdles, and finally, spam and security concerns.  All these must be addressed before ubiquitous computing becomes feasible.  I also will respond to pessimist who do not believe the technology will ever be powerful enough.

Current mobile devices use power quite vociferously, especially when connected to the network.  In the future, personal devices must always be connected to the Internet, but they must also act as sophisticated recording devices, and they must also power audio output, and screens or retina projectors.  All these require a substantial amount of power, and such a device is partially infeasible today simply because the battery would have to e enormous and incredibly heavy.  Advancements in battery technology to produce lighter and higher-capacity batteries is promising, but a breakthrough in the field is needed before this problem is solved.  Interesting research appears in nuclear-powered batteries, and research techniques are getting closer to producing a prototype which might produce a charge for up to ten years (8).

Then, further work is required in the field of human-computer interaction.  Current cellphone keypads are tiny, primitive, and are extremely difficult to use.  Although it has become popular to write and read novels on cellphones in Japan (9), interacting with the web and your personal device using a cellphone keyboard will never be easy enough to allow for ubiquitous computing.  Voice recognition has not proven very accurate or useful, and will probably only be used in the future by transcribers and other niche users.  Presumably, we will continue to speak English, and use a full-alphabet keyboard to communicate with our personal devices.  That keyboard probably will not be a physical one, and more likely a virtual on-screen keyboard, but more work and research needs to be done in this area to support ubiquitous computing.

Another hurdle we must over come is software.  The current state of software engineering is in a woeful. Our computers are slow, insecure, and error-prone.  Fortunately, software is purely theoretical and bound only by human ingenuity, so I think careful planning and thought, careful observance the art of programming and user interface design principles, as well as study of human-computer interaction, will allow us to engineer programs and software architectures that allow ubiquitous computing to work well.

Another problem we face is with protecting the end-user from, well, other people.  Those people may be malicious crackers bent on breaking security and penetrating defenses, or they may be willful and persistent advertisers who must tell you about the latest and greatest "growth enhancers".  Ubiquitous computing will not be at all ubiquitous if users are constantly bombarded and bothered by spam and security threats.  Secure software to protect against attacks and viruses is possible to engineer, but preventing spam is a much difficult problem, mostly because it includes serious social aspects.  Technology is merely a platform for this age-old social problem to grow on.  Solving the issue of spam is vital to the success of ubiquitous computing.

Finally, some consider the technological requirements of such systems to be extremely prohibitive.  Hard drives are considered too small, cellphones too big and clumsy, and connections to slow.  These predictions promise to halt the dream of ubiquitous computing in its tracks.  However, if we consider the history of technological innovation in hardware, we encounter a surprising rate of growth.  In February 2005, Ray Kurzweil presented compiled data that show that this rate of growth is not merely fast, but also exponential (7).  Kurzweil collected several lists of technological growth in CPU power, Hard Drive size, memory and chip density, broadband speed, and other important indicators from various authoritative leaders in their fields, and compiled all this data to show that growth in all of these technologies is exponential.  Because of this amazing exponential growth, Kurzweil predicted in his 2005 book, "The Singularity is Near", that by the 2010’s we would have small computers integrated into our personal lives, access to high quality broadband connections, and small projectors in our glasses that would project an image directly on our retina.  If anything, the software to make Ubiquitous Computing will not be ready by the time the hardware exists.  Perhaps by 2015, the technology will mature enough to become used by everyone.  Predictions past this point are foolish.

Ubiquitous Computing promises a technological future that is extremely appealing.  Several trends are already pointing towards the functionality we will see, and there is ample evidence that the growth in hardware innovation and broadband Internet and cellular connectivity will more than support the infrastructural needs of ubiquitous computing.  There are several problems that still stand in the way of ubiquitous computing, but these hopefully will be resolved by the research efforts working in these fields.  Ubiquitous Computing at first sounds like technology from the far future, but when considering past technological growth, we see that ubiquitous computing is actually a mild and tame prediction, probably manifesting by 2015, and the more radical changes will actually appear only in 20 or 30 years.

Works Cited 

[4­]: Caverly, Doug. "California Earthquake Prompts Twitter Reaction." webpronews.com 31 Oct. 2007. 12 Dec. 2007
<http://www.webpronews.com/topnews/2007/10/31/california-earthquake-prompts-twitter-reaction>.

"ICT Statistics." International Telecommunications Union. International Telecommunications Union. 13 Dec. 2007 <http://www.itu.int/ITU-D/ict/statistics/ict/index.html>.

[1]: Jan Chipchase: Our Cell Phones, Ourselves. Jan Chipchase, Nokia Research. 2007. TED Talks. 12 Dec. 2007 <http://www.ted.com/index.php/talks/view/id/190>.

[5]: Messina, Chris. "Twitter Hashtags for Emergency Coordination and Disaster Relief." FactoryCity. 22 Oct. 2007. 13 Dec. 2007 <http://factoryjoe.com/blog/2007/10/22/twitter-hashtags-for-emergency-coordination-and-disaster-relief/>.

[6]: "MyLifeBits Project - Microsoft BARC Media Presence Group." Microsoft Research. Microsoft Research. 13 Dec. 2007 <http://research.microsoft.com/barc/mediapresence/MyLifeBits.aspx>.

[8]: "New "Nuclear Battery" Runs 10 Years, 10 Times More Powerful." Physorg.Com 12 Mar. 2005. 13 Dec. 2007 <http://www.physorg.com/news4081.html>.