Act Beyond Borders

Act Beyond Borders

Environmental deterioration, poverty, and aging populations, health and medical services are all examples of contemporary challenges that are complicated, diverse, and constantly changing. To address such challenges, we must move freely among scientific disciplines and develop entirely new ideas. On the 20th anniversary of the founding of Sony Computer Science Laboratories (Sony CSL), we proposed a new scientific methodology for this purpose: Open Systems Science.

Open Systems Science is a way to study the real world, where every system is open-unlike the discrete, clearly delineated “closed systems” of conventional science. The fundamental structure and condition of an open system can never be fully understood, and its behavior is difficult to predict. The only way to solve problems that arise from the properties of such a system is to get out of the research lab and see the reality.

At Sony CSL we are working for the future of humanity. We do not confine ourselves to basic research. Rather, our mission is to make a positive contribution to the world we live in. A laboratory must be more than a place where people conduct studies in pursuit of “the truth.” In our view, research that is academically advanced has value only if it advances an understanding of reality in all its complexity-and this very complexity requires us to constantly reappraise the systems that we study. Our motto, “Act Beyond Borders,” reminds us to be intellectually agile and to reconsider each topic from many angles. This mindset enables us to open up entirely new domains of research, and achieve major breakthroughs that lead to new businesses and industries.

This is no easy task. It demands a clear vision of the future and an unshakable commitment to a better world. It requires the rigor and courage to acknowledge any gap between what we want to see and what we actually see, and a resolute determination to bridge that gap. It is an endeavor that involves tackling countless challenges. But each challenge we overcome moves us a step closer to new knowledge that will contribute to a better world. To approach a vision of the future, we continue to do “Act Beyond Borders.”

Hiroaki Kitano
President & CEO, Director of Research
Sony Computer Science Laboratories, Inc.

Open System Science

Open Systems Science

In Sony CSL’s early years, researchers used a conventional scientific approach to study artificial intelligence and large distributed systems, but something about this struck us as not quite right. Eventually it became clear to us that the source of the problem was what we were studying: open systems.

A conventional scientific approach is to isolate a domain of study, analyze its constituents, and work out how the various constituents fit together to make the whole. But these days it is not easy to isolate a study domain, let alone break it down into its constituent elements. The domains where knowledge now needs to advance quickly are characterized by their tendency to influence, and be influenced by, other domains. Examples include the environment, sustainability, economic phenomena, and life itself. These are “open systems.” In each of these open systems, numerous factors interact in complicated ways, and subsystems with the potential to change each larger system are in constant flux.

Open systems may be manmade, such as the vast digital systems and man-machine interactions that are made possible by the internet. Huge systems of this kind have constantly shifting boundaries and service requirements. As a result, it is virtually impossible to know enough about any specific component of the system. Moreover, for such a system to be truly user friendly, a great deal needs to be known about the people who interact with it-but human beings, too, are extremely multifaceted, and their behavior is strongly influenced by the context in which they operate. In short, it is impossible to learn anything of significant value about any open system by attempting to reduce it to its constituent parts. A conventional scientific approach simply does not work.
Sony CSL therefore devised a new methodology called Open Systems Science, and we have been using this methodology since 2008. Open Systems Science retains the existing methodological tools of the scientist-analysis and synthesis-and introduces a new one: management. The methodology of Open Systems Science was subsequently made explicit as follows.

  • (1) Define the problem and its domain.
  • (2) Construct a model of the problem domain in detail based on first principles, making every effort to leave out nothing significant.
  • (3) Apply the model and see if it develops any inconsistency or contradicts the behavior of the system as time passes.
  • (4) If it does, revise the model or devise a new model. Expand, reduce, or change the problem domain if necessary.
  • (5) Repeat until a satisfactory result is obtained.

Conventional scientific methodology leads to an ever more detailed understanding of an ever more narrowly defined topic of study. Open Systems Science, on the other hand, offers a way to solve a problem in its real-world context, including how it influences and is influenced by different but related problems. This is precisely why we need an “Act Beyond Borders” approach. Lab studies may deepen a researcher’s understanding of an isolated phenomenon, but you cannot understand that phenomenon in relation to others unless you study it in its living, changing, real-world context. In fact it is no exaggeration to say that without such a mindset, Open Systems Science would be impossible.

Mario Tokoro
Founder
Sony Computer Science Laboratories, Inc.
Spring, 2015
Go to top