Program

Monday, January 29

8:00 Coffee

9:00 Welcome and Conference Goals
Suzi Iacono, NSF & Brian Kahin, Michigan

9:15 Keynote: Stuart Feldman, IBM

9:45 Infrastructure for Knowledge and Innovation
Dick Burk, OMB, chair
Peter Freeman, NSF [paper] [slides]
Brett Frischmann, Loyola University of Chicago [slides]
Maria Pluvia Zuniga, OECD [slides]
Steve Jackson, Michigan [slides] [report]

11:00 Break

11:15 Designing the Virtual Organisation
Susan Turnbull, GSA, chair
Paul Avery, Open Science Grid [slides]
Mark Ellisman, NCMIR
Michael Tiemann, Red Hat [slides]
Chris Mackie, Mellon Foundation

12:30 Lunch

1:15 Technology-Enabled Knowledge
Donald A.B. Lindberg, National Library of Medicine, chair
Larry Prusak, Babson
Dominique Foray, EPFL
Marla Capozzi, McKinsey
David Lightfoot, NSF [paper] [slides]

2:30 Keynote: Arden Bement, National Science Foundation [speech]

3:00 Break

3:15 The Ecology and Design of “Open”
Elliot Schwartz, Committee for Economic Development, chair
Elliot Maxwell, eMaxwell & Associates/CED/Johns Hopkins [slides]
Rishab Ghosh, MERIT/Maastricht [slides]
Joel West, San Jose State [paper] [slides]
Andrew Wyckoff, OECD [paper] [slides]
JC Herz, DoD/OSD

4:45 Break

5:00 Between Public and Private: Bridges, Fences, and New Terrain
Stephen Wolff, Cisco Systems, chair
Dan Burk, Minnesota [slides]
Fiona Murray/Kyle Jensen, MIT [slides]
Siobhán O'Mahoney, Harvard [paper] [slides]
Robert Graybill, Council on Competitiveness [slides]

6:15 Reception hosted by Science Commons

Tuesday, January 30

7:45 Coffee

8:15 Keynote: Congressman David Wu [speech]

8:45 Pooling and Integration
Kaye Husbands, NSF, chair
Gavin Clarkson, Michigan [slides]
Joseph Scott Miller, Lewis & Clark
Geertrui Van Overwalle, Katholieke Universiteit Leuven [slides]
Carl Cargill, Sun Microsystems [slides]
Shane Greenstein, Northwestern [slides]

10:15 Break

10:30 Architecting the Knowledge Commons
Karen Skinner, NIH, chair
Sara Boettiger, PIPRA [slides]
Brian Athey, Michigan
Arti Rai, Duke [slides]
John Wilbanks, Science Commons

12:00 Lunch

12:45 Standards Development under Pressure
Cita Furlani, NIST, chair
Andrew Updegrove, Gesmer Updegrove/consortiuminfo.org [slides]
Linda Garcia, Georgetown [slides]
Timothy Simcoe, Toronto [slides]
Eric Miller, formerly W3C/Dublin Core [slides]

2:00 Break

2:15 Aligning Patents and Knowledge
Ron Marchant, UK Patent Office, chair
Michael Meurer, Boston U.
Richard Jefferson, CAMBIA
Mario Biagioli, Harvard [paper]
James Toupin, U.S. Patent & Trademark Office
Sean Dennehey, UK Patent Office

3:15 Break

4:00 Policy Roundtable
Brian Kahin, University of Michigan/CCIA, chair
Jim Turner, House Science Committee
Cita Furlani, NIST
Ron Marchant, UK Patent Office
Dan Atkins, National Science Foundation

5:15 Adjourn

Themes

Cyberinfrastructure enables the creation, use, reuse, combination, organization, and sharing of knowledge within a virtually integrated environment of mixed resources.  In contrast to the Internet, an all-purpose globally accessible platform, cyberinfrastructure is constructed around specific projects, research communities, or unique resources.  However, it is not limited by its origins.  It can support technology transfer, inter-sector collaboration, public education, even commercial ventures. Full realization and optimization of cyberinfrastructure requires understanding its organizational, economic, and legal context.  The design of cyberinfrastructure may influence business decisions; it may unbalance or rebalance laws and policies.  Often, perhaps increasingly, the relationship is indirect, mediated by different forms of public and private ordering. 

The virtual organization and architecture of knowledge

While initially focused on stable research communities, cyberinfrastructure can support knowledge creation and management whether the task is organized on the basis of privileged access, property rights, contracts, frameworks for exchange, or simple commonality of interest.  It can surmount traditional barriers and define new boundaries loosely or sharply, instantly or permanently. It can enable ad hoc groups to focus on particular tasks, lines of inquiry, or commonality of purpose with unprecedented efficiency, flexibility, and speed.  It can subdivide, focus, fork, spin off, extend, expand, and contract as circumstance and opportunity suggest.  The inherent capability and potential of cyberinfrastructure suggests a wide range of strategic opportunities – and possible vulnerabilities.   At the same time, it offers tools to mitigate high transaction costs, promote openness and transparency, and help address information, market, and policy failures. 

Cyberinfrastructure has evolved out of IT research community where it serves in variety of roles from infrastructure to laboratory to research subject – and offers a potentially pathway to private sector innovation.  It is also of special importance to biotechnology, especially bioinformatics, because of the close relationship between science and technology and their mutual dependence on common research tools and databases.  The design of cyberinfrastructure in these two critical areas needs to take into account the special practices and institutional architecture of knowledge, technology transfer, and innovation.

Measuring knowledge, cyberinfrastructure, and collaborative innovation

The Web has democratized access to information, but the costs of managing and sharing knowledge remain high, especially when human evaluation and interpretation is called for.  Managing knowledge in research is a continuing, costly process that extends from gathering new data to organizing, evaluating, reevaluating, and diffusion.  It extends into practical innovation -- application, product development, marketing, and support.  The cyberinfrastructure vision contemplates the virtual integration of these processes to improve the cycle of research, learning, and teaching in higher education and improve technology transfer, innovation, and commercialization throughout the economy. 

Despite the potential value of cyberinfrastructure, it remains difficult to define and monitor key components and complements.  Knowledge, collaboration, and innovation are all complex and context-dependent, as distinct from readily identifiable units such as bits, individuals, and patents.  Even units that appear meaningful, such as licenses, prove ambiguous and multivalent under close inspection.  The very ability of cyberinfrastructure to breach and bridge boundaries contributes to the problem of creating meaningful statistics.  The success of cyberinfrastructure in reducing information and transaction costs increases opportunity for free exchange and reliance on complementary sources of value, undermining the significance of priced markets as economic indicators.

Nonetheless advances are being made in measuring economic value indirectly to show the value of standards, software, databases, and other forms of public intellectual property.

Normative, institutional, and contractual models for open innovation

It is important to distinguish two different aspects of open innovation, both of which are supported by cyberinfrastructure: 1.) open resources in support of research and innovation; and 2.) open processes in research, innovation, and production.  In both cases, “open” is a matter of degree.  The degree is easy to see for resources in terms of cost, access, and restrictions.  In processes, it is matter of who can participate and the procedures for making decisions.   While there is often a synergistic relationship between the two, the distinction needs to be understood.  For example, traditional standards organizations have been scrupulously concerned with process, less so with access to the resulting standards.

Open resources

As innovation becomes more complex and cumulative – and more data-dependent, knowledge-intensive, science-based, and user-oriented – organizations becomes increasingly dependent on external information, knowledge, specialized tools, and other resources.  Infrastructure facilitates moving information and knowledge in and out organizations, increasing the potential for combining capabilities and resources in new or inexpensive ways.

In particular, infrastructure may facilitate the use of common resources to leverage investments in complementary assets.  When do research organizations choose to invest in developing shared resources?  How are shared resources organized and supported?

Open processes

The development of open standards and open source software are the most institutionalized, widespread, and well-known processes of open innovation.  Both are enabled by the Internet-based infrastructure.  Both are relatively public, although they are more controlled and hierarchical than is often assumed.  These are organized under principles of reciprocity and transparency, as reflected in common understanding on how intellectual property is treated.

There are also a variety of more private, proprietary forms of open innovation, in which firms look to suppliers, customers, and partners to help evolve and refine products and services.  With the exception of certain alliances, these are typically asymmetrical, based on relationships of complementary interests reinforced by intellectual property and contract.

Relationship of cyberinfrastructure to open resources and open processes

Experience with the Internet, the Web, and related applications demonstrates powerful relationships among open resources, open processes, and infrastructure in information technology, especially software.  This interaction appears dependent on inherent intimacy of resource and process in software as well as the ability to allocate and structure tasks to fit readily specified outcomes.  The relationship between knowledge and cyberinfrastructure is less direct in other fields, but it may be related to factors such as importance of shared data, semantic enhancement of datasets, dependence of technology on science, open advertisement of research in progress, publishing practices, university-industry interaction, professional and disciplinary networks, licensing and pooling of patent rights, incentives to secrecy, etc.  Developing and maintaining software involves very different incentives, costs, and activities than maintaining a shared database, and the supporting infrastructure must be tailored accordingly.

Managing incentives for standards development

IT standards development has become the paradigm of collaboration in creating new markets, although it has evolved out of a long history in other technologies where standards served a number of other purposes.  More than other outcomes of collaborative innovation, IT standards are essential to enabling collaborative innovation itself.  Their success demonstrates consensus on the importance of cooperation in promoting innovation and competition.  The success of the Internet and the Web in particular demonstrates the importance the power of nonproprietary standards as platforms for rapid uptake and innovation.  However, increasingly ambitious IT standards have come into growing conflict with an expanding patent system and incentives to hide information and exploit knowledge assymetries and failures.  Today, there is concern that different aspects of standards development – technology choices, patent information, licensing terms – should be pragmatically aligned and integrated in order to increase business certainty and accelerate standards development, adoption, and implementation. 

Commercialization and Privatization

Cyberinfrastructure is subject to change over time in many different ways.  The user base grows in size and diversity.  Shared resources evolve and mature.  New needs and opportunities are identified.  While it is unlikely to go commercial in the way that the Internet opened up to widespread and unconstrained public use, cyberinfrastructure faces issues of how access and participation by commercial users should be allowed or encouraged, along with proprietary resources and functions as elements within cyberinfrastructure.  On the one hand, growing diversity of participants is desirable, especially if there is the promise of new products or services.  But diversity also can also create tensions and, more insidiously, opportunistic behavior.

The Internet was built on well-defined well-understood platform of nonproprietary technology that was built on top of private lines and was easily extended to commercial users based on principles of free and open interconnection.  Cyberinfrastructure is a vision of software-integrated access and process that has no specific technological embodiment, but consists of a shifting variety of heterogeneous resources with different formulations of ownership, control, and access.  While privatization of the Internet was framed by principles of telecommunications policy, privatization of cyberinfrastructure is more an ad hoc matter of individual intellectual property strategies than public policy.  Whereas commercialization and privatization of the Internet followed a classical downstream trajectory of technology transfer, knowledge management and process integration are presently just as critical and technology-dependent within firms as within research communities.  Nonetheless, there are broad differences in the scope and degree of sharing knowledge across borders and the far more public and transparent nature of cyberinfrastructure. 

Incentives and knowledge

Patents and other forms of intellectual property are designed to encourage the creation and disclosure of information and knowledge that might otherwise be kept secret.  The strong controls that patents provide allow firms to share proprietary knowledge with suppliers, customers, partners, and collaborators – increasing opportunities for external relationships and diminishing the need to keep innovation within the firm.  While awarded patents are by definition public, applying for patents is private – completely so for the first 18 months, a long time in the IT sector.  Patenting behavior can undermine collaboration on knowledge and innovation, and obscured intellectual property interests can jeopardize the essential purpose of cyberinfrastructure – the ability to use and reuse information and knowledge.

Private ordering solutions

There are a number of solutions to perceived problems of opacity, overpropertization, coordination, hold-up, and proprietary capture – all of which work against the capacity of cyberinfrastructure to manage knowledge in a coherent and straightforward manner.  These solutions include patent pools, cross-licenses, non-assertion agreements and covenants, declaratory public licenses (Creative Commons), and the varieties of open source licensing.  To what extent are these solutions used successfully to reduce uncertainty and costs of networking information and knowledge?

Public policy solutions

Often the need for confidence in understanding the nature and scope of rights in information and knowledge has been overshadowed by the interest in providing incentives and lessening burdens on rights – as in the abandoning of notice and registration requirements under copyright.  But sometimes logical mediating procedures are introduced, such as the notice-and-takedown provisions of the Digital Millenium Copyright Act.  The U.S. Patent and Trademark Office is experimenting with the use of peer review and differentiated examination procedures to better allocate the costs of knowledge and build confidence in the quality of issued patents.  Reform legislation has proposed changes in the standard of willful infringement, which presently discourages researchers from looking at patents.

For more information, please contact Brian Kahin at kahin@umich.edu.