Let's look back to the organizational principle underlying AIST. "Organizations are possible only when you have people." Independent researchers gather at their own initiative around unit heads, who are independent thinkers, thereby forming a research unit. Even though the unit has taken shape in an autonomous way, it demonstrates an organizational principle unique to AIST. Specifically, one research unit has three different types of researchers-those engaged in Type-I basic research, those performing Type-II basic research, and those involved in product realization research-and these researchers collaborate coherently and concurrently in their research pursuits. We refer to this organizational principle as Full Research. Backed by this principle, each of our units has demonstrated fluid progress in their research endeavors. That research has shown steady gains not only in formal quantitative accomplishments, but also in quality, whether measured in terms of ingenuity or in contributions to industry.

Suffice it to say that Full Research has taken root at AIST. Of course, depending on the research unit, that may at times be only partially the case. This is particularly true for the Research Center, which faces project deadlines. However, from a long-term perspective, this is not really an exception to the rule because individual units can be understood to be making progress on their own, discrete elements of a larger undertaking in Full Research. Individual units at AIST are each engaged in their own ventures in Full Research. However, in terms of working together toward the achievement of a single research goal as pursued on an institute-wide scale, their efforts taken together form the structure of Full Research at AIST on the whole. The primary goal of that Full Research is to generate technologies essential to a shift by Japanese domestic industry to assume the leading role in the creation of sustainable societies on this planet.

Many topics were covered by the some 50 workshops held to date on themes relating to Full Research. Those workshops were spread across four series: Type-II basic research, the products of Full Research, Type-I basic research for Full Research, and strategies for Full Research. At each of those venues, young researchers, research unit heads, vice-presidents, and many staff from the administrative sector had opportunities to freely present their views and opinions. Given that its activities span such a broad range of research fields, AIST was, as to be expected, initially concerned about the communication problems participants might face. However, as the series of workshops proceeded, the seeds of discourse and dialogue rapidly took root, and from that point forward, researchers in different fields acquired the common language they needed to share in AIST's objectives. This "language" has become an invaluable asset to AIST.

Currently, AIST has 10 research coordinators who either individually or in pairs handle one of seven different research fields: life science and technology; information technology; environment and energy; nanotechnology, materials, and manufacturing; social infrastructure (geology) and marine science; social infrastructure (standards); and computer science. Research coordinators typically possess a wealth of research experience and solid track records of accomplishment as well as a comprehensive perspective that extends beyond their own field of expertise. As indicated earlier, on an organization-wide basis, individual projects at AIST add up to undertakings in Full Research. Independent research units are one of the fundamental characteristics of AIST. As such, they tend to be an assemblage of research teams driven by disparate goals. It follows that special efforts in coordination are necessary to ensure that all research units are integrated into Full Research toward a single, overriding goal. The greater the level of autonomy that individual research units possess, the more difficult integration of this kind is likely to be. Nonetheless, research coordinators at AIST are successful in carrying through with this exceptionally difficult task of structuring Full Research together at AIST without impeding the autonomy of its individual research units in any way.

The work of individual research coordinators entails preparing research strategies, budget allocation plans, and researcher employment plans for the fields under their supervision, through dialogue with the research units affected. The research coordinators' conferences strive for integration aimed at structuring together unit-level research operations into Full Research undertakings at the institute level. At the same time, through research reviews, they provide individual research units with feedback that conveys the goal of Full Research institute-wide. This involves much more than just management and communications. It is a process that continually provides researchers in a given field with guideposts for the pursuit of specialized research. To that end, coordinators develop insights not only into the latest cutting-edge research developments, but also into developments are likely to be made soon; predict the impact those themes may have on research in general and on industry; and put together a scientifically persuasive research roadmap that is based on their insights and predictions. And the resulting accomplishments of this specialized research are utilized at AIST for the formation of innovative theories relevant to the research field. On the whole, these approaches should be evident from AIST's recently completed research strategy.

The coordinators' conferences have also been announcing interesting results, including the fusion of nanotechnology with biotechnology, and the proposal of new industries through collaboration between the fields of measurement and standards technology and other fields. These conference results do not just impact research, but also are generating mechanisms for developing researchers equipped with entirely new specialties. Through the amalgamation of an even broader spectrum of research fields, the seeds for "AIST-industry-academia projects" are being borne.

As discussed in the preceding section, research coordination has the benefit of carrying individual research unit coherency and concurrency to an institute-wide level. This benefit, however, can also be expected to extend beyond AIST's confines.

The Science and Technology Basic Law of 1995 as well as the subsequent Science and Technology Basic Plan include provisions that strongly advocate cooperation between industry and academia. This is a global trend, as underscored by the "Science in Society and Science for Society" segment of the Declaration on Science and the Use of Scientific Knowledge issued at the 1999 World Conference on Science (in Budapest). Various events have been held in Japan, including Business-Academia Collaboration Summits. Cooperation between industry and academia has become an important goal, together with the incorporation of Japan's national universities. That is because close cooperation by academia and the business community is understood to be a realistic and effective way of returning public funds invested in basic research under the Science and Technology Basic Law back to society and individual citizens. However, this is not to imply we have a clear understanding of the forms of business-academia cooperation that are best suited to this goal, or how best to realize the cooperation.

According to FY 2002 data, research expenditures, including personnel costs, totaled 16.7 trillion yen that year. Of that total, 3.5 trillion yen was supplied by the government, mostly through funding allocated to universities and public research institutions. The rest, or 13.2 trillion yen, came from the private sector, and almost all of that was utilized by private institutions. These flows of research funds do not intersect. Except for research funding programs that have been set up exclusively to finance extremely short-term undertakings in joint research, there is no incentive for business-academia cooperation in the research domain. Overcoming this impasse will demand the introduction of a public research funding framework that encourages collaboration by the business community and academic sector (similar to the Intelligent Manufacturing System (IMS) international joint research set up by the Ministry of Economy, Trade, and Industry). In addition, though, it will be necessary to devise mechanisms that guarantee that cooperation brings beneficial results. To this end, AIST has begun work on the following structures.

The structures are designed to spread Full Research beyond the confines of AIST. Full Research comprises the coherent and concurrent pursuit of three different types of research-Type-I basic research, Type-II basic research, and product realization research-by a single research unit, thereby leading to the formation of a process in which the achievements of basic research generate social value. An individual research unit needs to comprise at least several dozen researchers for this process to be implemented in complete form. At AIST, this is feasible because each research unit employs 52 (full-time) researchers on average. However, the situation at universities is different. Based on available statistics, Japan has a total of about 54,000 working professors. Including doctoral students, the number of researchers under a university professor's supervision averages about five. This accordingly suggests that to engage in Full Research undertakings, a university would have to create a special provisional organization for the purpose. That organization, moreover, would have to allow its member researchers to collaborate in a coherent and concurrent fashion; functioning purely for liaison purposes would be useless.

Within the university setting, organizations are usually created with an educational focus. For example, university departments are composed of professors with different research interests within a given academic field who join together to offer students a balanced curriculum. Accordingly, the professors who come together to form the faculty of a department have created an organization for the ultimate purpose of providing education; research is led by individual professors and cooperation with other professors in that research is not always possible. For this reason, research led by faculty members within the university setting is usually conducted for the purpose of receiving acclaim from academic societies in the same field. And in most cases, it is limited almost entirely to Type-I basic research. Generally speaking, universities may be described as organizations ideally suited for undertakings in Type-I basic research. In terms of fostering academic progress, this form of organization does not present any inconsistencies. Still, universities run into trouble when they seek social value for research accomplishments in keeping with the mission of a particular academic field. Engineering is a classic example.

When striving to create social value with the fruits of basic research, research papers written in the interest of earning acclaim from academic societies typically fall short. In our experience, this goal usually requires that Type-II basic research or product realization research also be performed. However, if that is not possible within the confines of the university, researchers then have the option of utilizing the intellectual property of their research accomplishments for joint research with private companies, or licensing it, or launching independent ventures of their own. That is certainly something to be encouraged. Nonetheless, the more original and novel the results of Type-I basic research, the more indispensable basic research in other areas-namely, Type-II basic research or product realization research-will be to the task of assigning those results real value. Furthermore, achieving balance with Type-I basic research will be a difficult undertaking for small groups of researchers. This reality has become an obstacle for many universities even in the U.S., a country known to be a world leader in promoting collaboration between business and academia. If university professors were to pour their resources too heavily into cooperation with industry or into their own private ventures while remaining members of a small-scale university research organization, they could risk compromising their educational obligations or their own highly original Type-I basic research within the university setting. It has been reported that the trend now is toward placing restrictions on such sideline ventures.

As a way to help avert difficulties of this nature, AIST is planning to open up Full Research to the outside. In effect, this will be collaboration between universities and AIST. People engaged in Type-I basic research in a university setting would engage in joint research with researchers pursuing Type-I basic research at AIST. That approach would lead naturally to a fusion with AIST-led Type-II basic research and product realization research. Private industry would then be the next step. From the university perspective, this approach would provide a bridge to collaboration with private business and help lower the costs associated with such collaboration. Although these costs would be essential to drawing corporate expectations and obtaining information for engineering research, they must be kept within reasonable limits. Furthermore, in their effort to lower these costs, if universities prove too eager to satisfy corporate expectations, they may lose sight of their fundamental research goals and compromise the originality and novelty of their research. Conversely, from AIST's perspective, cooperation is possible as long as the university is engaged in Type-I basic research on common themes.

And, needless to say, for the collaborative bridges to be worthwhile from the university perspective, Full Research at AIST must lead appropriately to industrial applications or benefits for society at large. That is the fundamental goal of Full Research and although I will not elaborate here, let me underscore the connections. Through Type-II basic research or product realization research, Full research by individual research units leads to connections with society in the broad sense, that is, through the products of research that each research unit shares with society. However, when collaboration is with private business in the narrower sense, practical approaches range from joint research and consigned research to licensing agreements, personnel transfers, personnel exchange, and the startup of venture companies. To facilitate these forms of collaboration, it is imperative that AIST, as a public institution that conducts mainly publicly funded research, have a transparent relationship with cooperating companies, which, as private institutions, are driven by the profit motive. For example, in joint research undertakings, relationships between research progress and injections of public research funding must be clearly defined. Also, problems will arise in determining the extent of public funding that can be provided for the launch of high-tech startups that will capitalize on the fruits of publicly funded research. These issues are difficult to quantify. However, to frame it in simpler terms, as noted earlier, the greater the originality or novelty of the accomplishments of Type-I basic research, and consequently, the stronger the impact those accomplishments are expected to have on private industry, the less likely collaboration will be. Heavy and sustained injections of public funding will conceivably still be necessary after the results of Type-I basic research have been obtained. In the case of high-tech startups, the accomplishments of basic research will presumably be at too basic a level to translate into anything worthwhile. However, if those accomplishments are considered to have value in facilitating a shift of industrial infrastructure toward sustainable industrial pursuits, sharing them with corporations would conceivably be meaningful as a way for a public institution to support startup ventures. Further, experience with that process would presumably provide knowledge of ways to convert the accomplishments of basic research into actual value. Accomplishments that can be expected to translate easily into successful venture startups need not have support from public institutions.

In this way, universities, AIST, and private industry could establish clear relationships with one another. In fact, AIST has already established cooperative relationships and begun cooperating with several universities on the basis of these concepts. Additionally, it has already entered into collaborative ventures with private corporations based on a variety of models, including comprehensive agreements with large companies, theme-specific agreements with smaller companies, and institute-wide plans for the promotion of high-tech startups. The AIST-industry-academia project I touched on earlier constitutes a crucial trial undertaking aimed at integrating these approaches into one.

AIST is now in the process of drafting plans for the nationwide creation of suitable schemes for collaboration with universities and private industry through a synthesis of these varied tie-up formats. Given that other public research institutions in addition to AIST are likely to possess comparable research frameworks, these models could conceivably be explored for purposes other than research on industrial technologies. As a device for facilitating the generation of social value from basic scientific research, this would constitute a network model for innovations that evolve into new Japanese inventions-in other words, a network of excellence. AIST is committed to serving as a "hub of innovation" within that network.

In the previous sections, I stated that we have already embarked on a path toward building a uniquely Japanese network for collaboration with universities and private industry by having our research coordinators extend Full Research to the institute-wide level, and from there, to the national level, while retaining independent research units as our foundation. That is the path AIST has explored and set for itself by applying its strengths in full. However, in the process, AIST has learned about something else: namely, problems associated with the methodologies for research institution management and operation.

Although it has been four years since its inauguration, AIST is still a young organization. Furthermore, given that it is one of the first independent administrative institutions of its kind-an organizational structure with which Japan does not yet have much experience-AIST still has some time to go before it can be described as mature. The management approach for independent administrative institutions involves the assignment of initial goals set by administrative organs to the institution and the achievement of the goals by the institution primarily with public funding, albeit with a certain degree of autonomy in operation. The institution is then evaluated for its performance, and is assigned new goals depending on the outcome of that evaluation. However, many hitherto unknown problems confronted this management approach. We are learning much as we solve those issues.

The management of AIST has focused on research autonomy which is fundamental to research institutions: in other words, the autonomy of its research units. However, responsibility for achieving the mission assigned AIST by national government institutions, that is, the execution of assigned objectives, is guaranteed by giving the president the power to alter or eliminate research units. In fact, over the past four years, many research units have been created, reorganized, or disbanded as AIST has pursued its research strategy. AIST has learned about the methods used for those purposes, and is in the process of codifying them. During this period, each research unit demonstrated major strides in the arena of Full Research. And in the process, AIST gradually put into tangible form the management and operation practices it has adopted for its research units. Those practices cover everything from hiring, personnel management, evaluations, and budget allocations to collaboration with other sectors, information gathering and dissemination, international cooperation, and intellectual property management. To date, AIST has continued with efforts to improve the environment for research by devoting consideration to numerous internal rules and regulations aimed at supporting research efforts. They had to do chiefly with the diversification of hiring methods; the design of career paths; the evaluation of research units and individual researchers, and the use of evaluation findings for management purposes; budget allocation formulas; intellectual property strategy; and procedures for the establishment of new ventures. It can be concluded that those efforts have helped AIST assume its proper "shape" as a research institute that is an independent administrative institution and design itself as a forum for research. Organizational design and operation are matters subject to the optimization- and efficiency-oriented goals that are assigned to independent administrative institutions. This, too, is something that AIST has continued to pursue day in and day out thanks to its officers’ strong awareness of AIST as a pioneering independent administrative institution. It is urgent that lessons learned in these areas be shared and quickly utilized. Although I do not have space to discuss each and every one here, I will consider presenting them for institute-wide discussion and debate at the earliest convenience. Here, a general overview will have to suffice.

Here it is my intention to discuss matters relating to the structure and sharing of fundamental operational roles involved in the performance of work, and to the clarification of the powers and responsibilities associated with the sharing of those roles, all of which are common for management and operation practices aimed at supporting research bearing the diverse content described above. Every job incorporates the following elements: the presentation of problems, analysis of the problems, drafting of plans of action, deliberation, enactment, enforcement, evaluation, revisions, and so on. The first requirement is that the structures formed by these elements become an essential loop for functional evolution. Looking back over the past four years and the revisions that were made each year to arrangements that had been established the year before for many research fields, it seems reasonable to conclude that this loop has operated properly. The formula employed for budget allocations is a classic example, and was something that researchers tended to view as ambiguous. However, conditions have begun to settle down and the organization at last seems to be exiting from the unsteady phase it experienced upon startup. Nonetheless, the loop I mentioned above has not always been visible, and this problem relates to matters I will discuss below.

A second requirement is the clear division of responsibilities and authority, but further study is needed here. In short, one may say the separation of power within the organization has not been clarified sufficiently. Although it may seem awkward to apply the concept of the separation of power, as taken from classic theories of the nation-state, to an institution with clear objectives like AIST, we are of the view that we are now being asked to deal with fundamental organizational theories that require precisely such concepts.

There is a fundamental relationship between formulation and enforcement. As noted earlier, AIST found it necessary over its first four years to draft numerous rules and regulations. These were then put into effect and enforced. In the process, though, there were concerns that the bounds of authority and responsibility dividing formulators, enactors, and enforcers had not been clearly defined. As a consequence, there were cases where work concentrated on certain sections, people knew not where to turn when they faced problems during implementation, and certain departments were at times subjected to undue criticism. Although all of these problems were resolved thanks to the diligent efforts of the responsible personnel, it cannot be denied that efficiency suffered in the meantime.

To better portray the parties concerned here in their traditional form, suffice it to say the Planning Headquarters was responsible for formulation, the Board of Directors (Executive Committee) for enactment, and the various vice-presidents for enforcement. Although the president should bear general responsibility for the institute, the powers and responsibilities of formulation, enactment, and enforcement should clearly revert to the three groupings mentioned above. Furthermore, it should go without saying that clarifying where authority and responsibility lie effectively encourages mutual dialogue by the holders of authority. Another aspect of the separation of power is evaluations and audits, which Montesquieu would have termed as judicial power. Currently within AIST, the independence of these duties is being cultivated.

Idealistic arguments aside, the prime objectives of management and operation are to establish a climate that facilitates the implementation of research which is our ultimate objective and to strive to do this in an efficient way. AIST has endeavored enormously to that end over the past four years. And, it has amassed a track record of accomplishment and learned its lessons sufficiently in the process. By doing its best to systematize and apply these, it is anticipated that AIST will move a step closer toward the attainment of its goals.

The recently enacted AIST Charter, "Science in Society, for Society," was prepared by the Drafting Committee which is composed of 10 young researchers headed by Reiko Azumi as Chairperson. This Charter has earned the overwhelming support of all AIST personnel, including senior researchers, and has even emotionally moved many of those who read it. Although that is directly attributable to the excellence of the Charter's content and written style at the hands of the Drafting Committee's young members, it is not the only reason. Indeed, it is my feeling that the Charter stands on the one hand as a declaration toward the future, and on the other, as an emotionally moving statement that succeeds in expressing with brevity the hardships shared by all who have led independent administrative institutions over the past four years. Today, we all stand on the threshold of a future that the second research phase will bring, having together shared in the lessons of those last four years. Further, as expressed in the AIST Charter, we can declare with confidence that AIST will adequately satisfy the commitments it has been assigned in Japan, which is moving forward with a Science and Technology Basic Plan in accordance with provisions of the Science and Technology Basic Law.

Japanese society, including Japanese industry, clearly has an instrumental role to fulfill in creating a sustainable society, now a challenge for all humankind. Elaborating on the scale and quality of this role will demand the creation of "products for society" that are derived from fresh scientific knowledge. And to that end, AIST is determined to move forward, into tomorrow.