An Interview with 2017 Marconi Award Winner Arun Netravali

An Interview with 2017 Marconi Award Winner Arun Netravali

Arun Netravali, the 2017 Marconi Fellow, was the ninth president of the legendary Bell Labs. In addition to his pioneering work in video compression based on motion estimation (which provided the base technology for developing HDTV, digital TV and a broad array of video services), he promoted a more collaborative approach to research at Bell Labs, beginning with his own HDTV/MPEG research team.

Dr. Netravali sat down with us to discuss his work and insights into Bell Labs’ extraordinary research accomplishments.

Q. You’re credited with “pioneering contributions to digital video technology, systems, and services.” What specific applications came out of your work?

AN: My pioneering work was on video compression based on motion estimation, description and compensation. This work led to recent digital video compression standards including MPEG 1, MPEG 2, MPEG 4, H.261, H.263, H.264 (MPEG 4 Part 10), H.265 (HEVC), VP8 and VP9, serving as the key base technology. It’s now used in commercial video broadcasts (digital TV, HDTV, and UHD TV), DVD’s, streaming services such as YouTube, and other digital video services including video messaging, video conferencing, and video communications, and in all digital television sets and mobile phones today. It is remarkable to see projects grow from their nascent stages into full-fledged commercial deployments.

Q. You spent nearly your whole career at Bell Labs. What did you do there and what did you enjoy most?

AN: At Bell Labs, I launched research in video coding and compression in the 1980s, and HDTV and video networking research in the 1990s. During this period, I championed big system initiatives like HDTV and softswitch. I enjoyed seeing how the results of our teams’ research influence lives – and the bottom line of our company (including HDTV, ATLANTA, PACKETSTAR, PATHSTAR and many more systems).

Q. Bell Labs has often been cited as one of the greatest research labs ever. What has made it so special?

AN: If I had to answer in a word, I would say “people.” It was (and, in a different way, still is) a special place because of the people who worked there. Bells Labs prided itself on hiring exceptionally talented people. And hiring was not the sole domain of HR. Line management devoted considerable time to cultivating relationships with universities and their faculty to understand who the top recruits might be. Recruiting, especially from university campuses and research labs, was a major process for technical management.

Once on board, new hires learned rapidly from others, not just technical aspects, but team behavior and collaboration. Management was selected not just for technical skill, but also for the ability to help develop others. It wasn’t perfect, but our goal and processes were designed to select, train, and retain the best. Most often, people came to work there not just to work for an organization, but to work with the people who were there.

People had a sense of purpose; regardless of one’s role, we all knew that the fundamental mission was excellence in communications. It began with voice communications, and during the 20th century expanded into the larger companion disciplines of data communications and computing. To provide the necessary technology one had to understand the fundamental science —and to do this required people trained in many disciplines and a willingness to help and mentor others.

Bells Labs was fortunate to be well-funded for most of the 20th century by parts of AT&T. Yet the environment was much more than buildings and equipment; it was a way of interacting with others. It was simultaneously egalitarian and a meritocracy. Though about a third of the Technical Staff had PhDs, the title was rarely, if ever, used internally. Many problems required interdisciplinary thinking and nearly all staff would take time to answer questions from a colleague. Labs and offices were arranged in ways to encourage interaction. Individuals were selected for leadership positions, either technical or management, because of merit or achievement. Whether it was individual achievements or major team accomplishments, Bell Labs was extraordinary because its people were extraordinary.

Q. Your colleagues have commented on the “magic” of collaboration in the development of HDTV and MPEG. How did this come about?

AN: I created one team, with all the disciplines required for an end-to-end solution, with one common goal – to advance the state-of-the-art in video. I wanted our great team to see how research could impact society. To make HDTV and MPEG “real”, we created a new type of organization where researchers and developers sat side-by-side to deliver new technologies quickly to the market. New research ideas in digital video – based on fundamental research advances – found their way quickly into deployed products. I applied this approach to other research areas at Bell Labs including packet networks, fiber optics and all scalable IP Cellular Networks.

Q. What inspired you to become a scientist?

AN: In some sense, I fell into a scientific career. I recall seeing one of my most studious friends working in the library shortly after we had finished our final exams at the end of junior year of high school in India. He was filling out an entrance application to attend college at Indian Institutes of Technology (IIT). Without even knowing what IIT was, I assumed that if my smart friend was doing it, I should do it too.

I passed the IIT entrance exam, and initially studied chemical engineering. Very quickly, I realized that the required “chemistry” portion of my studies involved plenty of memorization, i.e., the Periodic Table of Elements, which I simply detested! I switched to electrical engineering and fell in love with the subject (even though I was still required to learn the periodic table.)

Q. How did your education influence your career?

AN: After graduating from IIT, I came to the US to attend Rice University as a graduate student. I am so grateful to the faculty and administration at Rice for sponsoring my education. Through the university, I had the pleasure of living with a host family, who not only provided me with accommodations, but also embraced me as a family member. In return for their generosity, I served as a companion and tutor for their grade school- aged children. I felt a huge sense of personal satisfaction in their academic achievements, and it was then that I realized the immeasurable value of mentorship, which I have paid forward in my own career.

During this time, I also worked as a short-order cook in a diner, as a taxi cab driver, and as a ball boy at a local country club. I had the chance to interact with individuals from different cultural and socioeconomic backgrounds, and truly understand that America’s strength comes from her diversity. When I graduated with my PhD in 1970, I not only had learned how to conceptualize and conduct research independently through my formal education – I had a strong desire to positively impact individuals in every corner of America’s melting pot.

Q. Who were your key mentors and inspirations?

AN: I’ve had so many! One important one was C. Chapin “Chape” Cutler, my first boss at Bell Labs. He taught me lessons with lifelong impact. I noticed he kept a leather notebook with him, and when I asked about it, he told me that he spent one hour every day listing and documenting new research ideas. I started to do this too, and over the years, I realized that my best ideas were first conceptualized in my notebook, which to this day is always in my pocket. Chape also insisted that we conduct meetings while walking up several flights of stairs because “only healthy minds can generate great ideas.” Even now, I rarely take elevators.

Q. What do you consider your single most important achievement?

AN: Two things: having had the opportunity to guide and enable the greatest minds at Bell Labs to generate their best work; and my children. My daughter completed an MD-PhD from Harvard Medical School/MIT. She is a physician-scientist focused on bridging the patient’s bedside with the scientist’s bench. She recently characterized the cellular pathway of development for an important immune cell, a question that had evaded scientists for over fifty years. My son is currently pursuing a PhD at MIT in Computer Science, focusing on computer networking and distributed systems. His recent projects aim to improve the performance and debugging aspects of large-scale Web services. I am so pleased that both have chosen careers in which professional success is measured by the ability to positively impact and help others.

Q. What challenges did you have to overcome during your career and what effect did they have on you?

AN: I had to stand up for what I believed in. A technology change is often a mind-set change, and this is hard for many people. There were many naysayers about the change from analog to digital video. I had to work to communicate long-term value, while mitigating transition overhead and risks.

Q. Did you ever have an “aha” moment?

AN: Yes. I noticed a big difference between natural scenes and video seen on analog TV versus what I saw in real life, and realized that it is because of noise and limited bandwidth. I believed that video should be delivered digitally to represent the real scene more closely. However, when digitized, video required so much bandwidth that it was almost impossible to deliver even a small video in real time. Then I realized that most parts of the scene rarely change from frame to frame. Because of that, video could be represented a lot more effectively if we could describe the motion of a few moving objects in the scene and send only differences to adjacent scene frames after compensating for such motion. Also, circuits could be packet-switched! That was a key insight in moving the project forward.

Q. What is your greatest regret?

AN: My greatest regret is not training and trying out for Wimbledon. Growing up in India, I was an avid badminton and cricket player, and when I first came to Rice University in Houston I was dismayed to find that neither was a mainstream sport. Instead, I begged the coach of the Rice Men’s College Division I Tennis Team to let me learn how to play tennis by practicing with the team members. Over the period of a year, I routinely beat their top singles player, and started to win several professional tournaments. Because I did not think that I was good enough to become the next “Roger Federer”, I pursued my other passion and followed the safer route to become a scientist. Even now, I wonder what might have been, had I had the courage to train and try out for Wimbledon.

Q. Anything else you want to say?

AN: I’m proud that the Marconi Prize recognizes not just scientific achievement, but science that provides practical benefits to humankind. The video compression technologies and digital HDTV system that my team developed have been commercialized by consumer electronics companies such as Zenith, RCA, NHK, Sony, Toshiba, Panasonic, Samsung Electronics, and LG Electronics.

The same video compression technology is employed in all smart phones – Android and iOS phones – for video streaming, video messaging, and video communications.  MPEG has also influenced all commercial video and DVD production – movies, YouTube, cameras, etc. – and I feel very good about that.