We are now used to thinking of computers and software as being ubiquitous and
as the engine that drives many aspects of our social and economic lives. Today,
most business activities and processes, manufacturing systems, personal
financial management, almost all types of personal and business correspondence
and communications, government processes, healthcare, and education are driven
by the power of computing and software. It goes far beyond the computing that
happens in a desktop computer, laptop or a server. Computing is pervasive and
present in a number of devices, including home appliances, automotive systems,
consumer electronic devices including the cheapest, and most humble of them all:
mobile phones.
The areas that Computer Science plays a critical role in are too numerous to
mention in detail, so I shall confine myself to talking about just a few in this
article.
Less than 10 years ago, a mobile phone was only owned by the privileged few
and the network coverage was spotty and unreliable. Today, as we all know, it is
carried by almost everyone who lives in urban areas and by many even in rural
areas. And a mobile phone is no longer merely a communications device. It is
already used widely for downloading media and other data, playing games,
exploring the knowledge-base contained on Internet and even for financial
transactions. The low cost, wide footprint, ease of use and convenient form
factor; all have already made the mobile phone the number one personal computing
device. Yet, we have barely begun to explore its true potential. To understand
this, consider the fact that each mobile phone also contains a number of
sensors; the microphone and GSM radio are sensors present in every mobile phone.
As one looks slightly above the entry level to more sophisticated mobile phones,
we find even more sensors like cameras, accelerometers, and GPS radios. This
opens up possibilities of creating a network of sensing and communicating
devices (ie, a 'sensor network') at a relatively low cost and without much
overhead for deployment, since the sensors are already out in the 'field.' Some
applications that immediately come to mind are traffic sensing, where the GSM
radio can be used to track the traveler's location, and the accelerometer can be
used to detect whether the traffic is flowing smoothly or is in a stop and go
scenario (typical of traffic jams). In fact, the accelerometer can even be used
to detect whether the road being navigated is bumpy or smooth. In developing
nations like India, where GPS radios are not common in mobile phones, the GSM
radio can be used as a good substitute, and can also offer the advantage of not
being constrained by needing line-of-sight to a satellite to be effective.
Dr. P. Anandan, Managing Director, Microsoft Research Labs India |
Healthcare is another important area where I see the mobile phone being
particularly effective. Given that large tracts of rural areas are typically not
well covered by the healthcare system or by wired telephony, mobile phones offer
a way to deliver quality healthcare. Research projects are already underway in
different parts of the world, that explore ideas such as Fetal Heart Rate and
Fetal Activity Monitoring systems that use the mobile phone network to track and
record fetal activity. Another project looks to use low cost microscopes in
conjunction with mobile phones in remote rural areas to capture and transmit
microscopic images wirelessly to any lab in the world for diagnosis.
We are so used to taking pictures using digital cameras and sharing and
viewing pictures online that it is easy to forget that less than 10 years ago,
most people were still taking pictures using film cameras and getting them
processed and printed at a local photo store. In fact there was serious
skepticism that digital cameras would ever produce pictures of the quality and
resolution to replace film cameras. Yet today, consumer level cameras produce
high quality pictures in rich colors that are over 4 mega pixels. We are at a
point where hundreds if not thousands of photographs of common scenes (such as a
famous monument or tourist site) are being taken every day covering every view
and aspect of the scene. Yet when we view these pictures later, we view them one
at a time, seeing only what happened in one instance in a small part of the
scene.
View from the Gopuram | Inner Courtyard Walk |
Inner Courtyard | Inner Courtyard with Video in the inset |
3D views of the Sri Andal Temple in SriVilliputtur, Tamil Nadu |
Technology is reaching a point where we can put these pictures together to
construct a whole 3D scene. Recently, Microsoft Research, in collaboration with
University of Washington introduced a new technology called 'phototourism' which
enables the construction of 3D scenes from photographs taken by the casual
tourist using their ordinary consumer level cameras. Today these technologies
take several hours to create the 3D model; however, with improvement in
technology and devices it is not hard to imagine these being done in a matter of
minutes if not faster. Other photo analysis technologies allow us to rapidly
capture many images of the scene with different exposure settings and create one
picture that contains the best view of the entire scene. These abilities suggest
ways of influencing how the picture is taken. One can imagine a scenario in
which the 3D model is provided within the camera viewer as a backdrop to the new
picture, thus allowing you to carefully position the camera to get the best view
of the scene. Or even rapidly show other pictures of the same scene that are
available so you can concentrate on filling in what is missing or improving the
ones taken before. And the viewing experience can move from that of looking at
isolated pictures one at a time to exploring the entire scene in 3D and in high
resolution. 3D views of the Sri Andal Temple in SriVilliputtur, Tamil Nadu are
shown on the next page.
It is a common mistake to equate the science of computing (namely 'Computer
Science') with programs and software. At the heart of this vast engine that is
powering the society and world economy lie a variety of powerful mathematical
principles and technologies such as algorithms, systems, architectures for
computing and communication, database principles and techniques, and programming
languages, environments, and methodologies. These fundamental aspects of
Computer Science are now extending their influence to fields outside computing
itself. It has begun to exert its influence in all sciences as an enabler in
their advancement. And this role as an enabler does not mean just providing
computers that speed up the process of scientific discovery, but participating
actively as a partner in the process of scientific research by bringing in its
own methodologies, discoveries and principles that can be applied to different
sciences. There are already a number of examples where Computer Science has
partnered with other sciences to bring about breakthroughs. Sciences such as
biology that focus on analyzing and modeling of proteins and molecules leverage
ideas from Computer Science such as pattern recognition and data mining and
machine learning in the very process of scientific discovery. In fact, pattern
recognition and machine learning, used extensively in areas such as spam
filtering (for email), are today being used to help in discovery of drugs for
deadly diseases like HIV. Computer Science is also being used extensively to
address another problem that threatens to become critical: hydrology. There are
a number of initiatives across the world that seek to leverage computer science
to study regional hydrology, collect, process, analyze and model information to
solve the problem of inadequate water.
Ultimately, it is the power of logical reasoning applied to empirical data
that leads to discoveries in science. And computer science, being the science of
logic as well as data manipulation and exploration is thus a natural partner in
the endeavor of scientific investigation.