PressCNRS international magazine

Table of contents

Joseph Sifakis

Keeping the World in Check

Both in and out of the lab, Joseph Sifakis' discoveries help decipher and master the very nature of information. His pioneering work on Model Checking, the most common method for verifying industrial applications, won him the 2007 A.M. Turing Award–considered the Nobel Prize for computer science.


© E. Perrin/CNRS Photothèque

Joseph Sifakis is the first French recipient of the Turing Award, the equivalent of the Nobel Prize for computer science.

His innovations may be solving the very real problems of industrialists around the world, but Joseph Sifakis (CNRS Silver Medal 2001 recipient) is also a philosopher. “In the middle ages, physics changed the way humans perceived the world. Who is to say that computer science couldn’t achieve the same thing today?” he asks with a smile. Born on the island of Crete, in Heraklion near the temple of Knossos, Sifakis earned a degree in electrical engineering from the University of Athens before moving to Grenoble in France for a PhD in Computer Science. As a student, he devoured the works of great physicists and spent long hours debating with his classmates the very nature of consciousness. Today, as founder of the Verimag Laboratory1 in Grenoble, he still brings to computer science the same breadth of mind, and the same ability to see the whole universe in a single particle, or the human soul in a computer chip. Skipping with ease from the highly theoretical to the very practical, he is just as active in the lab as he is in recruiting industrial partners to fund and develop his work. Indeed it is for both his fundamental and applied contributions that he has been awarded the A.M. Turing Award.
The prestigious prize, granted by the Association for Computing Machinery (ACM)2 and widely considered the equivalent of the Nobel Prize for computer science, is shared this year by Sifakis and a team of American scientists: Edmund M. Clarke of Carnegie Mellon University and E. Allen Emerson of the University of Texas at Austin.
Working in competition since the eighties, all three researchers pioneered what is called Model Checking. This verification technology enables computer hardware and software engineers to find errors efficiently in complex system designs. Logical errors in these systems, ubiquitous in cars, airplanes, and industrial plants, as well as in telecommunication technologies and e-commerce security protocols, often result in delays in getting products to market. Model Checking provides an algorithmic means of determining whether the design satisfies a formal specification, such as a pattern for a sequence of events, and identifies the counter-examples that show the source of the problem. In other words, it provides greater assurance that the systems will perform as intended by the designers.
Today, many major hardware and software companies rely heavily on Model Checking thanks in no small part to Sifakis’ involvement in the field. He was particularly instrumental in getting industry leaders on board by co-founding the first International Conference on Computer Aided Verification (CAV).
For the last ten years, though, Sifakis has wished to take computer science to a level beyond verification and the trial-and-error method by which computer systems are usually created. For this tireless researcher, who compares his profession to a vocation, fundamental research into computer science remains a vastly unexplored territory. Some believe that, because computer systems are incredibly complex, fundamental laws of behavior and predictability cannot be derived from them. Yet Sifakis foresees a more constructive future for the discipline. By exploring the interface of information and matter, he hopes to find new ways of creating more reliable, more autonomous–and in the end–more intelligent machines. His current research activities include component-based design, modeling, and analysis of real-time systems that focus on correct-by-construction techniques. This means elaborating complex systems whose behavior would be predictable before they are even constructed. “I want to get closer to the physics paradigm, where to build a bridge, you start with a law and then pile on the bricks accordingly,” he explains.
Indeed, as computers pervade society, much is left to understand of the nebulous notion of information, which is just as important as matter and energy in the shaping of our daily lives. For Sifakis, who in his spare time also studies the evolution of languages, particularly ancient Greek, understanding how concepts and consciousness emerge is the first step to making machines more intelligent. Could the answers to these age-old questions really be found in a computer science lab? Can the language of computer science refresh our understanding of human nature? Capable of both successfully developing a technology that meets the needs of industry, and of placing abstract thought under a microscope, Joseph Sifakis has only begun to explore this exciting new discipline.

Lucille Hagège

Notes :

1. CNRS / Université Grenoble 1 / Groupe Grenoble INP

Contacts :

Joseph Sifakis
VERIMAG, Grenoble.


Back to homepageContactcredits