South Korea's scheme of launching network-based robots gained a boost

Network Robot Project Gets Boost

Carnegie Mellon’s Raj Reddy Manifests Much-Touted 80/20 Rule

By Kim Tae-gyu Staff Reporter South Korea's scheme of launching network-based robots gained a boost after a world-famous artificial intelligence (AI) expert confirmed Korea is heading in the right direction.

Carnegie Mellon University (CMU) professor Raj Reddy, one of the most respected scholars in the promising AI field, made the point in an e-mail interview with The Korea Times.

``Yes, this (the network-based robot) is a fine idea. Absolutely, the 67-year-old said.

The Ministry of Information and Communications plans to kick off feasibility tests of five different network-based robots _ three for home usage and two for post offices _ from late next year.

The wheeled robots' sensing and processing functions will be operated based on the nation's state-of-the-art broadband network while the machines provide mobility to the capability foundation.

However, Reddy cautioned possible perils of the mobile dummies since they can pose a serious threat to their owners as imaginatively explored in Spiderman II.

``However, distributed robotics is a complicated problem. Many checks and balances are needed to ensure mobile robots do not get out of control, he said.

On the development path of robots and AI, Reddy has provided an uncanny insight to the world, clearly manifest in the much-touted 80/20 rule.

It refers to the concept of making the computer perform 80 percent of the task while leaving the other 20 percent to the human being.

The rule has the advantage of speeding things up. If a problem is going to take 20 years to solve, probably about 80 percent of it can be solved in 10 years.

Reddy's basic idea is that it would be better to use other 10 years to settle 80 percent of other problem rather than grappling with the remaining 20 percent of the original one.

He predicts the rule will help accelerate the development of bona fide automatic robots, slashing the task of human owners to as low as 1 percent in a century.

``In each cycle, 20 percent of the 80/20 is a continuously changing number. Hundred years from now, 1 percent still must to be done by humans, the India-born visionary noted.

Asked to evaluate Korea's IT industry, he was positive, saying: ``IT industry is broad and multi-faceted. Korea has decided to invest in selected areas and in most of these are doing very well.

But, he added, there are other areas where Korea does not have a significance presence, and it should work on deploying more resources and people into those areas.

In fact, Reddy has contributed much to Korea's fast development in high-tech areas like software industry. He also helped many Korean outfits contact various CMU programs and learn from them.

In recognition of such efforts, the Korean government gave him a Bronze Tower award in 2002. And early this year, the Information and Communications University, the newly established IT education institute, named him an honorary professor.

Man on a Mission

Raj Reddy is an illustrious visionary in the artificial intelligence field and is now serving as professor in the School of Computer Science at Carnegie Mellon University (CMU).

He was born in India and received his master's degree in civil engineering in 1958 from the University of Madras in the world's second-most populous country.

He then moved to Australia and obtained a second master's degree two years later from the University of New South Wales in Sydney and worked as an IBM salesman.

For more study, Reddy crossed the Pacific to Stanford University, the birthplace of Silicon Valley, and gained his doctorate in computer science at the prestigious university in 1966.

After securing a Ph.D, Reddy started his academic career at the university as assistant professor and has been a member of the Carnegie Mellon faculty since 1969.

He served as founding director of the Robotics Institute from 1979 to 1991 and also worked as dean of the School of Computer Science over the following nine years.

During the period, many other prominent researchers had gone on to greener pastures but Reddy continued his commitment to the AI and poured his time and energy into it.

Such devotion was recognized in a big way when the U.S. government appointed Reddy to co-chair the President's Information Technology Advisory Committee in 1999. He helped set up the nation's high-tech policies at the position through 2001.

In 1994, he received the Turing Award, the highest honor in computer science.

The 67-year-old also has sustained special connections with South Korea and was tapped as an honorary professor of Information and Communications University in March, 2004.

The following is a question and answer session with professor Raj Reddy. - ED.

The Korea Times: You have contributed much to the development of Korea’s software industry and helped many Korean outfits contact various Carnegie Mellon University programs. Why have you done these things for Korea?

Raj Reddy: Korea has always been a progressive, hard working and disciplined country to become a model of Asia and we can learn how to continually re-apply the ``lessons learned’’ of Korea's success.

We have a dynamic team of Carnegie Mellon people, led by John Kang, and this dedicated capability keeps Carnegie Mellon connected about Korea.

KT: You are an honorary professor of Korea’s Information and Communications University (ICU). What do you think of Korea’s IT talents?

Reddy: In the sphere of IT and communication, Korea has been a leader in hardware and communication networking, IP broadband and networking, etc. In other aspects of IT (e.g., software development for embedded systems) Korea is strong, but not yet a world leader.

ICU, by its name, is responsible for both these kinds of technology, and is working with CMU for understanding of how Korea can provide leadership in all areas over the long term. ICU has recruited its faculty and students from the best sources, and I am impressed by the quality of the faculty and students.

KT: Will we see the appearance of ``a Raj Reddy of Korea?’’

Reddy: There are many distinguished Korean scientists worldwide, including our Peter Lee in Carnegie Mellon, with an internationally preeminent future.

KT: Please evaluate Korea’s IT industry or infrastructure.

Reddy: Korea's IT industry is clearly strong. Its communications infrastructure is the envy of the world. This includes broadband, and wireless networks and applications for all kinds of e-business, e-government, and so on.

This provides a foundation for further development of new ideas. However, other countries are trying to catch up, so Korea should continue to move forward with continuous new development.

KT: What is the nation’s strength and what is its drawback?

Reddy: Korea's main strength is the people. They are intelligent, disciplined and hard working. Perhaps the main drawback is the language barrier.

KT: Is Korea’s IT industry heading in the right direction?

Reddy: The IT industry is broad and multi-faceted. Korea has decided to invest in selected areas, and most of these are doing very well. There are other areas where they do not have a significant presence, but it is just a question of how best to deploy resources and a small population of 48 million.

KT: You have contributed much to American IT policies by serving the President’s Information Technology Advisory Committee (PITAC). What policies are you most proud of?

Reddy: The PITAC worked to identify key issues of broad importance in the United States, including education, health care, e-governance, e-business, e-security, digital libraries and the digital divide as potentially important areas for new investment and research.

By proposing sets of action plans in these areas, the PITAC tried to make its recommendations to maintain national attention on these areas.

Is there something you regret among things you decided during your tenure at the PITAC?

Toward the end of my tenure, we produced a fork of identity theft recommendation. But for bureaucratic reasons, there was no publicly released. However, the problem areas of ID theft must be a serious focus of our society.

KT: In your own career, you started and continued emphasis on robotics and artificial intelligence (AI), even though many left these fields for greener pastures. Why did you maintain your emphasis on these fields?

Reddy: I continue to work in AI and Robotics as my primary intellectual activity and scientific activity. In doing so, I followed examples of Carnegie Mellon's thought leaders, the intellectual giants such as Perlis, Newell and Simon.

They encouraged young faculty to explore further development of the applications of computer science as a set of great challenges. Many other activities are related to societal issues and public policy. They are equally important and need to be pursued by scientists, engineers, executives, and everyone who can contribute to such endeavor.

KT: What are the promises and challenges of artificial intelligence?

Reddy: We will have super human capability to improve our capability. The challenge is to build systems that can learn from experience, and operate in human real time.

KT: Do you still think the 80-20 rule is the right direction for robotics?

Reddy: The 80/20 rule is usually meant that 80 percent of the value comes from only 20 percent of the most important activities.

However, in the instance of robotics, at any given time, we wish to find ways to automate more and more. It suggests that 80 percent of current activities can be done by robots, and the remaining 20 percent are done by human beings, under a continuous optimization problem.

In each cycle, 20 percent of the 80/20 is a continuously changing number. Hundred years from now, 1 percent still must to be done by humans.

KT: Korea is developing network-based robots. The government plans most of sensing and processing functions from the broadband network, while the robot adds mobility to this capability foundation. From your perspective, is Korea doing the right thing?

Reddy: Yes, this is a fine idea. Absolutely. However, distributed robotics is a complicated problem. Many checks and balances are needed to ensure mobile robots do not go out of control, as we see in the ideas of Spiderman II.

KT: How will intelligent computers (like the tera PC and peta PC) affect the way we live and work?

Reddy: We can apply increasing computation capabilities to our knowledge activities, such as learning healthcare, many other functions of private and public sector activities, and the many requirements of home life.

Creating these capabilities will demand significant computational power connected to sensors, such as imaging sensors, and actuators, to perform the needed functions.

The computational power of the tera PC, provided that it is affordable, will significantly enhance human productivity.

KT: Are there any other suggestions that you would like to communicate to Korea?

Reddy: I continue to be inspired by the Korean people, and their accomplishments. In this rapidly changing technological world, we need to develop mechanisms for the continuous advancement of human resources.

We must continue to explore the usefulness of various models of education, such as intelligent tutoring, learning by doing, and just-in-time learning.

We need to continuously enhance our ability to use technology and methods, so that the best ideas are continually developed into routine applications, to help society to move forward.