eScience Institute - IT Connect

Download Report

Transcript eScience Institute - IT Connect

Conversations with University of
Washington Research Leaders
Ed Lazowska, eScience Institute
William Washington, LST
Janice Fournier, LST
Tom Lewis, LST
Erik Lundberg, eScience Institute
eScience Institute
Project Goals
Learn about future directions of research
Understand role of technology in research
Identify resources and services that researchers
need
eScience Institute
Project Overview
 Partners: UW Technology and eScience Institute
 First large-scale assessment of researchers’
technology needs conducted at UW
 UW is first among its peer institutions to implement a
project of this type
 50+ technology professionals involved in project
eScience Institute
Selection Criteria
 Number and monetary amount of current grants relative
to others in similar disciplines
 Prestigious recognition by career status:
 Junior faculty: Sloan Research Fellowships, Packard
Fellowships, NSF CAREER Awards
 Senior faculty: National Academies membership
 Peer recommendations
eScience Institute
eScience Institute
Participants
 Phase I (July 2007 - July 2008)
 38 researchers interviewed
 Phase II (Summer 2008)
 89 researchers interviewed
 Total:
 127 researchers interviewed
 264 UW researchers contacted
 48% response rate
eScience Institute
Data & Analysis
 Phase II data: audio recordings, field notes, interview
summaries
 Identified unique categories of needs
 Noted both unmet needs and solutions
 Prioritized based on number of unmet needs
eScience Institute
Findings
eScience Institute
eScience Institute
Research Climate
 Funding constraints
 Voluminous data
 Interdisciplinary and inter-institutional collaborations
eScience Institute
eScience Institute
IT Climate
 Rely on UW’s advanced networking infrastructure
 Leverage external resources
eScience Institute
eScience Institute
Researchers’ IT Needs:
 IT & Data Management Expertise
 Data Management Infrastructure
 Computing Power
 Communication & Collaboration
 Data Analysis & Collection
 Additional Resources
eScience Institute
 IT and Data Management Expertise
 Local technology support—inadequate access
 Data management expertise—designing new
systems, enhancing current practices
 Information—technologies and expertise available at
UW
eScience Institute
eScience Institute
 Data Management Infrastructure
 Access to storage infrastructure—large quantities
of data for current projects and data archives
 Data backup—inconsistent systems among
researchers, some unreliable practices
 Data security—secure access needed for interinstitutional partners
eScience Institute
eScience Institute
 Computing Power
 Computing power—ever-increasing need for more
powerful machines
 Managing and housing computing clusters—
challenge to configure and house
 Network access—need for high bandwidth
eScience Institute
eScience Institute
 Communication and Collaboration
 Real-time collaboration technologies—critical to
some researchers
 Videoconferencing, web conferencing, and
advanced teleconferencing
 Everyday technologies—basic tools are used by
most researchers
 Phone and email, wikis and Web sites,
remote desktop access
eScience Institute
eScience Institute
 Data Collection and Analysis
 Analysis—need for specialized expertise
 Visualization—not yet widespread
 Collection—Web access to data, mobile devices
eScience Institute
eScience Institute
 Additional Resources
 Labs & equipment—need for communal resources
 Educational technology—strong link between
research and education
 Centralized information—support for research
administration
 Group pricing—negotiating discounts for products
and services
eScience Institute
eScience Institute
Recommendations
 For central institutional units—UW Technology,
eScience Institute, Office of Research, Office of
Information Management
 For the UW research community
eScience Institute
Assumptions
 Researchers depend heavily upon UW’s leadership in
networking infrastructure, and the UW must continue to
sustain and advance these resources.
 The UW should also pursue cloud-sourced solutions
whenever suitable.
eScience Institute
For Central Institutional Units
eScience Institute
eScience Institute
A New Data Management Paradigm
 Integrate human expertise with the hardware
 Data schema design, database design, parallel
computing tasks, data analysis, data mining
 Secure data management infrastructure options: backups, short- and long-term, high availability, cloudsourced
eScience Institute
On-Demand IT Expertise & Consulting
 Data management
 Security
 Network, cluster, and storage design and administration
 General IT support
eScience Institute
Greater Availability of Computing Power
 Sustained and dedicated
 On-demand, pay-as-you go
 High performance networks
eScience Institute
Communication & Collaboration Toolset
 A few, basic enhancements…
 Drop-in videoconferencing
 Ubiquitous Web-conferencing
 Cloud-sourced applications to ease access
 OpenID and federated AuthN/AuthZ
eScience Institute
Research Commons
 Bring together researchers, their support staff, and
central units
 Share information about what is already available
 Identify collaborative opportunities wrt. Projects,
technologies, facilities, support strategies
 Mechanisms for group pricing
eScience Institute
UW Research Community
 Work with us to build a culture of collaboration
eScience Institute
Questions/Discussion
 Conversations report will be available at:
http://www.washington.edu/lst/
http://escience.washington.edu/
eScience Institute
eScience Institute