ACKNOWLEDGEMENTS

As Duke’s Chief Information Officer, I charged a series of working groups in 2016 with drafting reports that would form the basis of a University-Wide Strategic Plan for Information Technology, with special emphasis on the technology infrastructure to support the wide range of local and global services. Seven university-wide IT working groups were constituted in order to develop a comprehensive IT plan for Duke that can support needs and ambitions of the broad range of university IT stakeholders. I want to thank the many individuals who led this effort and the more than 50 people who participated in the development and review of the working group reports and the current plan for their insights, energy, expertise, and investment in this most important strategic planning process.

Tracy Futhey

Vice President of Information Technology and Chief Information Officer

June 2017

Information Technology Strategic Planning Leadership

Working Group Chairs / Co-Chairs:

Linda Franzoni, Pratt / Lynne O’Brien, Library

Rachel Richesson, Nursing / Brenda Franks, OIT

Jeff Chase, A&S Computer Science / Charles Kneifel, OIT

Stephen Bass, A&S Physics / Kristen Brown, Public Affairs

Tom Nechyba, Sanford / Mark DeLong, OIT

Terry Oas, School of Medicine / Paula Batton, OIT

 

ITAC Strategic Planning Steering Group:

John Board (Chair) Associate CIO and Professor, Pratt School of Engineering

Timothy McGeary (Co-Chair), Associate University Librarian for IT, Duke Libraries

Joanne Van Tuyl, ITAC Chair, and Professor of Slavic and Eurasian Studies

Ginny Cake, Assistant CIO OIT

Charles Kneifel, Sr. Technical Director, OIT

Julian Lombardi, Assistant Vice President, OIT

 

OIT Senior Leadership:

Tracy Futhey, VPIT and CIO

John Board, Associate CIO

Richard Biever, Chief Information Security Officer

Julian Lombardi, Assistant Vice President

Charles Kneifel, Sr. Technical Director

Robert Johnson, Senior Director of Communications Infrastructure and Global Strategies

Christopher Meyer, Senior Director, Enterprise Systems & Support

Jennifer Vizas, Director, Projects, Service Adoption & Engagement

 

INTRODUCTION

This is a time of rapid change in information technology.  The rate of this change, both technological and cultural, is having a dramatic impact on university life and the academic enterprise.  Since the writing of Duke’s last strategic plan in 2005, there have been many significant technical advances – most notably across the following areas:

• Networking
• Virtualization
• Security
• Mobile computing
• Collaboration
• Cloud computing
• Research computing

Beyond these, there have also been strong cultural shifts in the acceptance/use of information technology across our university community.  Information technology is now a ubiquitous presence in the lives of most of our community members as smartphone-enabled mobile computing has gained traction in the consumer arena.  Since 2005, the mobile computing revolution has put a highly capable networked computer in most everyone’s hands nearly all the time – along with national advertising campaigns extolling the virtues of such devices across our personal and professional lives. The resulting emergence of powerful and easy-to-use applications for doing any manner of things online at any time has shifted the larger culture - along with our university community’s expectations about information technologies and the services they enable.  Information technology is now so integral to modern university life that IT organizations must act more as partners within the academic enterprise rather than as only the providers of secure and reliable IT infrastructures and services.  Close partnerships with those who advance Duke’s academic mission are necessary to ensure our agility, efficiency, and competitiveness as a leading educational institution at a time of very rapid change in information technology and in its impact on the evolution of educational and research practices.

Environmental factors considered in this report include:

• Continuing acceleration in the pace of change in information technology
• Increasing complexity of technology services and their integration
• Service needs of a globally-distributed university community
• Changes in the way that faculty and students use technology and access knowledge

• Security threats that are ever-changing and increasing
• Institutional goals and priorities of the 2017 Duke University Strategic Plan

In Duke’s relatively decentralized IT environment, individual schools and other administrative divisions often maintain their own local IT support organizations that partner with OIT. In these partnerships, OIT typically provides technical and infrastructure support while the local units provide specific functional support and direction. This approach to IT support is designed to enable greater innovation within the distributed units since the local technical support organizations are able to use their resources to experiment with new services and technologies that fit the evolving needs of their particular school, college, or department. In developing this strategic IT plan for Duke as a whole, representatives from across all of our distributed IT environment came together to imagine a shared and coordinated future: a robust University-wide IT partnership and support structure – one that anticipates and meets the needs of our entire community.

PLANNING PROCESS

This planning cycle represents the third concerted effort that the University has undertaken to produce a strategic planning report for information technology.  In 1996, a task force was constituted through the Information Technology Advisory Council (ITAC), to produce both tactical and strategic recommendations.  In 2005, OIT led the development of an IT Strategic Plan that has informed our work to the present day.

The 2017 Duke University Strategic Plan recognizes the increasingly important role that IT infrastructure and services play in making possible a modern and globally integrated university community.  The larger and more encompassing plan underscores the importance of information technology as a pervasive theme throughout all of its major goals rather than as an individual goal within the plan. In response, the present 2017 Duke University IT Strategic Plan recommendations represent a synthesis of relevant inputs from across an array of university stakeholders (schools, university-wide working groups, and various administrative units). The scope of consideration for the present plan encompasses all campus-based IT infrastructures and services over a timeframe of the next five to ten years.

Our process involved establishing seven IT strategic planning working groups which contributed strategic recommendations across the following areas (see appendices I-VII for the full working group reports):

• Living and Learning (LL)
• Research Computing Support (RC)
• Communications and Infrastructure (CI)
• Administrative and Business Systems (ABS)
• IT Security (ITS)
• Support Models, Procurement and Licensing (SMPL)
• Online Presence, Web and Mobile (OPWM)

All recommendations from the working group reports were consolidated into an overarching set of top-level recommendations and implementation cost estimates by the IT Strategic Planning Steering Group.  That group also ensured appropriate alignment between the current IT plan’s recommendations and what is needed to support objectives articulated under the three main goals of the 2017 Duke University Strategic Plan (see Table of key alignments on next page).

	Duke University Strategic Priorities 2017-2021
University Goal 1: Empower the Duke faculty as scholars and leaders of the university’s intellectual communities	University Goal 2: Provide a transformative educational experience for all students	University Goal 3: Strengthen Duke’s capacity to address challenges—both locally and globally	University Goal 4: Create a supportive environment for research, learning and academic communities
Areas articulated in Goal 1 and addressed in this IT plan: Modernizing the science and computational infrastructure. to support scholarship, teaching and engagement. Enhancing and simplifying faculty access to shared cyberinfrastructure facilities to meet the needs of researchers and improve Duke’s ability to recruit and retain new faculty. Catalyzing student groups and research and learning communities. Providing seed funding for pilot research programs and incentives for group collaborations and community building. Maintaining a modern computational infrastructure to support Duke’s quantitative initiatives and emphasis on the sciences. Expanding programs to help integrate technological innovation into education and research.	Areas articulated in Goal 2 and addressed in this IT plan: Promoting the development of learning communities and building technological capabilities into informal communal spaces Providing undergraduates an equal opportunity to access and benefit from the best of Duke. Helping to expand opportunities for graduate and professional school students to transfer their education into an increasing array of career options. Encouraging students to sample intellectual areas outside of their comfort zones.	Areas articulated in Goal 3 and addressed in this IT plan: Maintaining a modern and secure IT infrastructure to protect network resources at the levels required and help “make Duke a hub for local-to-global problem solving and two-way engagement with external stakeholders”.	Areas articulated in Goal 4 and addressed in this IT plan: Creating a supportive environment for research, learning and academic communities. Creating ‘collision’ spaces on campus that facilitate community. Building toward the goal of having all classrooms support 21st century pedagogies, with customizable space and standard digital technologies.

STRATEGIC GOALS:

2017-2021

The IT Strategic Planning Steering Group reviewed all seven of the Strategic Planning Working Group recommendations along with a close consideration of the key objectives outlined in the 2017 Duke University Strategic Plan and identified a set of three top-level goals and general areas of effort for the 2017-2021 planning period.

Goal 1:  Help people use technology more effectively

A plethora of IT systems and services already pervade Duke’s educational, research, and administrative activities.  Duke operates many of these services directly, others are operated by external partners over which we have varying degrees of control; still other services are identified and engaged directly by the community members we serve with no involvement from or notice to our various information technology support organizations.  Our community members access these services through a broad range of devices, including traditional desktop computers, notebooks, tablets, and especially phones and other mobile devices.  Classrooms and other learning spaces are equipped with a diverse range of devices.  Despite the complexity of this ecosystem, our users expect a seamless experience as they adopt new services and transition between systems and devices.  Making our content and services fully accessible and readily useable to community members with vastly differing abilities is a related challenge and opportunity for Duke, as it is for all of higher education.  Furthermore, since the internet exposes Duke and all of our community members to constant information security threats to both institutional and personal data, the countermeasures we must deploy to mitigate such threats often run counter to enabling a smooth and seamless user experience.  Therefore, to balance these concerns while reducing user frustration, we must assume a posture of minimizing the inconvenience of necessary security measures throughout all of our services while also educating our community about why such measures are needed.

Helping people use technology more effectively and improving the user experience will require progress in a number of areas. We should:

1.1 Enhance user interface design and user experience across deployed systems[1]

It is fair to criticize many of our past IT systems as having been designed by technically inclined people with little or no formal training in user interface and usability design. Too often, details of user interaction are left as a last minute cleanup task after the “real work” of achieving the desired technical objective is met.  The field of user experience design has emerged in recent years to invert this process and promote the user’s interaction with the service as a design principle that is as important as the technical task the system is accomplishing rather than as an inconvenient afterthought.  Duke’s central and distributed units, have already begun to develop needed expertise in the “UI/UX” (user interface / user experience) field, and we expect that user centered design and end user friendly approaches to IT systems implementation will increase greatly over the planning horizon.  We need to enhance user interface design and user experience across deployed systems by incorporating modern principles of user-centered design via a campus-wide coordinated approach (mobile standards definition, APIs, etc.). To do so, we will need to:

1.1.1 Establish, communicate, and maintain UI/UX standards to promote a common, intuitive experience for all official Duke systems and services

1.1.2 Improve the overall user interfaces and user experiences of all IT systems and services with particular attention to our older systems

1.1.3 Improve integration of services so users can move between related systems without distraction or complication

1.1.4 Establish modern “responsive design” approaches across all web products so that web or other computer applications automatically adapt their interfaces to accommodate the size and capabilities of the particular device being used

1.2 Improve back-end integration of tools and systems[2]

Historically, our enterprise IT systems have interacted via regular but infrequent (often nightly) reporting processes in which one system writes out all information it is told might be relevant to another system, and then the other system consumes that block of information. In recent years, Duke has made significant advances in modernizing our approach to where different IT systems communicate information in real time via Application Programming Interfaces (APIs).  In addition to keeping our systems better synchronized with up-to-date information, APIs also permit immediate deployment of unanticipated, but appropriately authorized, sharing of data between applications that had not communicated before.  Use of API’s (in authorized ways) allows us to better support the data requirements of user-developed applications such as student-written mobile apps.  In this way, developers are able to access institutional data quickly, safely, and securely.  Though many of our newer systems already embrace this approach, it is essential that we now improve back-end integration of tools and systems to promote seamless access to relevant information from a variety of tools including those supporting older systems.  We must therefore:

1.2.1 Use modern APIs from all newly developed or acquired systems

1.2.2 Develop a comprehensive catalogue of institutional IT services and data resources

1.2.3 Establish data governance plans for all shareable institutional data

1.3 Improve web support for university units[3]

The availability of local web development resources varies dramatically across the institution. Some units possess the technical and/or financial capacity to maintain modern, useful, useable, current websites, while others struggle to keep even basic sites updated, secured, and on-line.  Local budget optimizations have caused some units to deploy lower-cost web solutions that have proven problematic from security and support perspectives.  This has led to higher institutional costs to remedy problems.  Our current central and unit web development resources are not sufficiently resourced to help realize this vision and will require augmentation. Given the criticality of highly functional and secure web presences to all aspects of a modern institution, and the importance that faculty generated web content has to the overall research enterprise (professional, lab, and workgroup websites etc.), we should establish an “entitlement” level of web presence for university units and enhanced unit support for deploying content via web/mobile via tiered web services offerings. We will therefore need to:

1.3.1 Identify and establish a base level of web “entitlement” services that units can depend on to provide functional, effective, secure, and maintainable web presences

1.3.2 Develop a tiered set of mobile application development services to help meet the needs of unit mobile development efforts

1.4 Improve use and support of collaborative tools[4]

We must build capacity and provide support for emerging forms of online teaching and learning. Since we are in a period of rapid innovation with regard to the nature of available IT tools and services, it is important for Duke to stay ahead of the curve by building capacity and providing support for emerging forms of online teaching and learning.

1.4.1 Provide enhanced local support in the schools and departments around video production, motion/animation/visualization, coding and course design including course “trailers”

1.4.2 Increase the availability and capacity of technology options to support online learning, including supporting open-ended experiments to help faculty explore new possibilities

1.4.3 Promote more effective uses of existing tools and services

1.4.4 Promote faculty and staff investigations into new and emerging IT tools and services

1.5 Improve technology in learning and meeting spaces[5] and for academic support systems[6]

The use and generation of media content is changing the way learning takes place in classrooms. Today’s classrooms are best configured as media-enabled learning spaces with simple to use controls for media capture and access to online content.  Duke’s classroom configurations vary significantly across the academic units; many rooms are already capable of recording and delivering excellent media experiences, while others are in need of reconfiguration and upgrade.  Common issues with many of Duke’s classrooms and meeting rooms include poor lighting, washed out projectors, lack of working audio, poor site lines, and lack of interface standards. We must improve classrooms and other learning and meeting spaces with enhanced ability to access online resources and capture and disseminate media via easy-to-use controls. Furthermore, we need to provide adequate support for physical and virtual environments that support learning outside of traditional classroom settings. This includes support for specialized recording and media capture facilities and spaces that support physical computing, makerspaces, motion capture, etc.).  Finally, it is essential that we modernize systems that enable academic support functions (student advising, etc.). We should therefore:

1.5.1 Establish media standards capable of supporting modern classroom and meeting space uses

1.5.2 Increase funding to accelerate classroom and meeting space renovations and technology upgrades across all units as a means of meeting the common standard described in 1.5.1

1.5.3 Establish seed funding to modernize and improve the systems that support academic services for both traditional on-campus students and for online students

1.6 Improve our ability to provide accessible versions of digital content[7]

All of higher education is cognizant of increasing societal and legal demands to make our media content available in formats accessible to persons with varying disabilities.  There are increasing indications that the historical approach of many institutions to meet these requests on an as-demanded basis will be considered inadequate.  There are also increasing signs that all media (videos, still images, audio recordings) will need to be produced in multiple accessible formats at creation time, and that this requirement may extend to all media produced by the institution for any purpose, and not just strictly classroom media.  We need to improve our ability to provide accessible versions of digital content to meet the needs of various differently abled constituents.  Duke will not be alone in facing this challenge as new expectations and mandates also become clearer, but it is prudent to expect that significantly expanded resources may be needed to meet this need, and the needs of compliance training for content developers over this planning horizon. In the meantime, we should:

1.6.1 Establish tools and services to support the automatic and economical generation of accessible digital content in multiple formats

Goal 2: Provide ubiquitous, secure services and systems

2.1 Maintain a modern protected network[8]

Our data network is critical to almost everything we do at Duke.  It has evolved significantly over the past planning cycle and has generally kept up with ever-increasing demands, despite the increasingly hostile external environment.  Duke has also embraced experimentation with evolving network technologies to keep us ahead of most of our peers in provisioning high capability network services.  Nonetheless, keeping our network capacities ahead of the demands of our users remains a formidable task, especially in view of the increasing security precautions we are compelled to take.  Maintaining the leadership position of our campus networking over the next planning horizon will require vigilance and progress in a number of areas.  We should:

2.1.1 Increase the bandwidth of our wired network

Increased wired network bandwidth will be needed, but not in all places at the same time and improved network monitoring tools will enable us to enhance the capacities of individual circuits in an economical way before they become overloaded.

2.1.2 Monitor and replace old campus fiber

Our original campus fiber optic network is approaching 30 years of age. We will see an increasing need to replace aging fiber and the conduits that house it.

Addressing this will involve a phased approach to refresh and deferred maintenance of our critical fiber infrastructure as it approaches the end of its useable life of 20-25 years. With installation beginning in the late 1990’s our fiber plant is now a mix of technologies dependent on what was available at the time.

2.1.3 Extend campus fiber

Funding associated with 2.1.2 does not accommodate future fiber initiatives in support of research networking, outdoor WiFi, and/or physical security connectivity.  All have become essential elements of our campus infrastructure.

2.1.4 Improve campus WiFi and cellular coverage

Though wired network connections are essential for our IT infrastructure, most of our users rely on their wireless connections (both WiFi and cellular) for routine interactions with our systems and the wider Internet.  We recently completed a major enhancement of cellular coverage inside all campus buildings, though this will require continual management as technology evolves (think 4G moving to 5G and beyond over this planning cycle).  Our campus WiFi environment faces coverage challenges in many areas of campus; recent changes in WiFi technology improve the bandwidth a well-connected user device can consume, but the radio signals needed by these systems do not penetrate the thick walls of many Duke buildings.  We have already conducted trials of denser deployments of WiFi access points in several Duke buildings to validate that we can restore excellent WiFi coverage across campus. Completion (and funding) of this new deployment is a major goal of this report.

2.1.5 Provide on-demand circuits for bandwidth intensive activities

OIT’s aggressive exploration of newly evolving networking technologies with grants from NSF and other sources in recent years positions us to already offer, perhaps uniquely among our peers, instantaneous deployment of dedicated very high speed networking circuits between arbitrary locations on campus.  Extension of this capability to locations off campus (at collaborating institutions, for instance) is now being tested and we anticipate that use of such circuits will become routine during this planning period.  Maintenance of this environment once our original NSF seed funding ends will require additional expenditures going forward.

2.1.6 Improve data management and data transfer options for researchers

Over the past several years we have seen a rapidly developing increase in the need for research-based data storage across the academic disciplines.  This trend requires that we establish support and policy structures to help us better meet our institutional data storage and preservation requirements and to address support for agency-mandated data access beyond active phases of research projects.   This will involve assessing the quantity, nature, and location of our digital research output, assessing the level of expected increase over the next 3-5 years, and establishing a set of institutional policies around long-term storage and preservation of digital research data.

2.1.7 Improve ease of campus/health system collaborations

The difficulty of sharing data across the boundary between Duke’s medical and campus sides has been a sore point for the institution for a number of years; indeed this issue was specifically called out for attention in the last strategic plan.  Though the situation has improved in the intervening decade, it is still far from seamless.  The campus and the health system maintain separate networks, supported by different technical groups, and enforcing very different sets of security and interoperability standards.  These differences come largely from the paramount need to protect the privacy of medical information on the health system side, and the historical approach was rather coarse, preventing even non-sensitive information from flowing easily across the boundary.  Recently, the two networking groups have worked to unify a number of their networks for campus-wide systems (telephones, security cameras), and the same approach can and should be extended to allow appropriate sharing of non-PHI (Protected Health Information) across the boundary in a much more transparent way.  Enhancements to the secure research network, described elsewhere, will provide enhanced opportunities to share various forms of sensitive data, including PHI, among collaborators in efficient and compliant ways.

2.1.8 Enhance connectivity and access for off-campus users

The RTP area is in the midst of efforts by several companies to provide gigabit network services to home users.  This is a welcome development that we expect many Duke users will take advantage of, however our on-campus endpoints for these remote connections will need to be enhanced to avoid becoming the new bottleneck for our off-campus users doing high-bandwidth work from off campus.  There is room for significant experimentation and innovation here, as this is a capability very few of our peers have had experience with.

2.1.9 Improve network management, reporting, and resolutions

Our network is a very complex piece of engineering, with a large number of interactions and interdependencies among all of its parts, encompassing not only traditional networking equipment but now also security devices, load balancing units, network management systems (NMSs such as Cartographer for instance), and others.  Our efforts are further complicated by the networks’ constant and necessary evolution.  Duke has made good progress in recent years by increasing the degree of automation we bring to bear in monitoring and diagnosing network performance.  However, this remains an active area of concern requiring additional effort in the coming years.

2.2 Respond to and anticipate evolving information security threats[9]

IT security has been the fastest growing part of our campus IT enterprise since the last planning cycle, and it is likely to remain one of the fastest growing in the present cycle.  Sadly, the creativity and energy of the attackers appears to be unbounded, with new and serious threats appearing with concerning frequency.  Much of our response to the threats in recent years has been to enhance capabilities within our central IT security group.  Today, our IT security group acts as a kind of SWAT team assisting both other OIT departments, and the schools and other units lacking dedicated security resources when IT security issues arise.  Though continued growth of our IT security group is anticipated, we should also work to move security expertise down into the various distributed IT units to further mitigate the growing threat. This can be achieved by some mix of enhancing the security skills of existing staff, and focusing some new hires on hybrid professionals who have both security and subject matter expertise.  We should do the following to address specific threat areas that are likely to require enhanced expertise and expense during the upcoming planning period:

2.2.1 Mitigate against exposures exacerbated by the ‘mobile-first’ strategy espoused elsewhere in this document

Implementation would be covered by 2.1.8 (see above).

2.2.2 Establish processes to ensure compliance across all systems and services in the face of ever-increasing regulations

2.2.3 Enhance security awareness/education for all users for both their own personal information, and especially for their handling of Duke’s sensitive data, when relevant

2.3 Explore moving key enterprise services to the cloud[10]

Duke has been cautious compared to some of our peers in embracing the “cloud” to meet critical institutional needs, and yet we have indeed employed cloud providers when a thorough analysis indicates the decision makes sense (our Office 365 email environment, and our Sakai learning management system, are two examples). We must explore the costs and benefits of moving key enterprise services to the cloud.  This will involve:

2.3.1 Explore the potential for cloud services in support of disaster recovery

One area where Duke could profitably explore enhanced use of cloud services is in the disaster recovery area. It is now plausible to imagine keeping a “freeze dried” version of most of Duke’s IT systems on standby at an external cloud provider, ready to spin up a shadow campus on short order.

2.3.2 Continue to evaluate the potential efficacy of cloud services when contracting with external providers

Dollar costs are an important consideration, but so are data security and privacy concerns, and the reliability of the entire path between Duke and the candidate provider.

2.4 Better integrate our IT purchasing processes with the needs of our users[11]

Duke has had challenges in communicating about the various technology purchasing options that are available to faculty and staff.  By improving communication, we can enable faculty and others to make more cost-effective and secure decisions when making IT expenditures. Duke purchasing units should work with major IT system vendors to:

2.4.1 Permit pre-configured machines to be shipped directly to end users with needed Duke software and hardware configurations already installed. 

In this way, machines would be fully functional right out of the box, obviating the need for configuration by local IT staff.

2.5 Improve ways to request support or consultation between IT units and OIT[12]

Duke’s central and distributed IT units work well together in many instances, however we are also still overly reliant at times on staff within a local unit knowing the “right person to call” to get something done rather than having fully transparent processes for our local units to get the help they need from OIT.  To provide a more effective way to request support or consultation between the central and distributed IT support units, OIT should:

2.5.1 Improve the process by which it responds to requests for support and consultation by staff from other campus IT units.

Implementation will involve better coordination between support units and a refocusing of existing staff efforts.

Goal 3: Support innovation in research and education

Duke has created a strong ecosystem to encourage innovation with emerging technologies both within and beyond the formal curriculum.  Efforts such as the Co-Lab and other enhanced lab offerings have given us valuable experience in supporting students and others in experimenting with new and emerging technologies. That said, we are currently working to better understand how best to support innovation and experimentation across the populations we serve.  As remarkably powerful computing devices and sensors continue to become more affordable and accessible to our users, especially with regard to the internet of things (IoT), we will need to be flexible and agile in our approach to supporting their expanding use across teaching, research, and other aspects of university life. Duke has lagged behind a number of our peers in fully leveraging the vast quantities of data that can now be captured campus IT operations.  Such data has great potential in helping us to improve the overall university experience and increase efficiencies across many of our systems and business units. We should work to position Duke as a leader in supporting innovation and embracing appropriate uses of emerging technologies and collected data. We must therefore:

3.1 Support the evolving computing needs of our researchers[13]

Beyond fast networking, Duke scholars and researchers need access to various kinds of raw computational horsepower and increasingly vast amounts of digital storage.  On the academic campus, Duke has recently made a commitment to meet the basic computational and storage needs of most of our research community with on-demand (and freely available) services, and the systems to do this have already been put into service.  We are also committed to providing cost-effective options to researchers with very high computational and/or storage requirements.  Proper support for research computing is multifaceted - requiring more than the provision of computation and storage.  In order to meet existing and emerging needs, progress is needed on a number of additional fronts.  We should:

3.1.1 Establish a process by which digital products of scholarship and teaching are appropriately secured and preserved.

Meeting the mandates of our funding agencies and the expectations of our community to preserve data and other digital research products for posterity requires new ways of thinking about data archiving.  Our library and IT organizations have already been deeply engaged in thinking about these problems, and viable solutions are at hand, though significant uncertainties around the rate of growth of these services could present funding challenges over this planning horizon.

3.1.2 Increase the reproducibility of complex computational research.

In collaboration with the health system, we have been investigating methods to preserve not just the raw data of an experiment or simulation, but the entire computational environment and chain of transformations used to turn the raw data into results.  Imagine that a specific calculation was done with version 4 of some package that only ran on a very old version of Windows; even though that same package may now be on version 9 on the latest version of Windows, to verify reproducibility, we would need to have a way to reconstitute that now-ancient version 4 environment and its old Windows system at a future date.  Despite significant technical challenges, we expect this to become a viable, useful, and perhaps necessary service over this planning horizon.

3.1.3 Improve and expand the protected data network and services to meet existing and emerging use cases.[14]

Duke has already created a protected research enclave, the Protected Research Data Network (PRDN), in which appropriately vetted collaborators can work with various kinds of sensitive data.  This enclave initially supported very basic analysis on modest computational servers for local users.  The environment continues to expand to support more software tools, greater and more diverse computational power, and external collaborators.  The approaches we have taken in the PRDN are likely to be good avenues to meet expanding security compliance requirements for various kinds of federally funded research, especially research funded by DoD.  We expect the PRDN and related services to expand greatly in this planning horizon.

3.1.4 Provide a secure computing environment accessible from personal devices. 

Though security is very difficult to guarantee on personally owned smartphones and the like, nonetheless our users are expecting to be able to manage aspects of their research data activities, even those involving sensitive data, from their mobile devices, making use of applications developed for this purpose. Since it will be a necessary for us to develop applications and mechanisms that permit these essentially insecure devices to nonetheless conduct safe transactions, we should focus on ensuring the security and privacy of Duke data consumed and used by these applications, especially ones developed internally.

3.2 Improve Duke’s competence in data analytics[15] related to institutional data

Duke has lagged many of our peers in our efforts to leverage institutional data for improving our educational outcomes and operational efficiencies.  To address this, we should:

3.2.1 Develop ways that we can better leverage the use of captured institutional data to further improve educational outcomes and computational research workflows. 

3.2.2 Provide convenient extracurricular education and training opportunities for Duke community members so that they may improve their levels of data science literacy.

In the same way that many of our students, faculty, and staff have realized the great potential of the technology and learned mobile app development outside of credit-bearing courses, we believe that there exists a demand for more co-curricular educational opportunities in data science literacy. 

3.3 Advance both curricular and extracurricular opportunities for IT education[16]

The IT literacy of Duke community members is important to our community’s ability to fully leverage the capabilities made possible by industry advances in IT.  These advances, along with the increasing specialization and complexity of IT tools and services drive the need for us to provide training opportunities to help all of our community members take full advantage of new and emerging capabilities.  While Duke provides for-credit courses to its registered students on virtually all aspects of modern information technology, that curriculum remains relatively inaccessible to a large part of our community.  We should enhance our efforts to provide specialized IT training and general IT educational opportunities to Duke faculty, staff, and students beyond Duke’s course catalogue as a form of continuing education in IT. We must:

3.3.1 Improve IT training opportunities around the use of advanced Duke deployed IT tools and services.

3.3.2 Improve co-curricular educational opportunities around emerging topics and key practices in information technology and data analytics.

3.4 Support and promote IT innovation[17]

Over the past few years we have been successful in promoting innovation in education, research, and outreach. From the Duke Digital Initiative to CoLab to OIT’s involvement with NSF funded research exploring new technologies to support our mission, stories of IT-supported innovation by students, faculty, and staff have become pleasantly routine over the past year.  We should build upon this progress by taking the following actions:

3.4.1 Enhance opportunities and support for early access and experimentation with emerging technologies.

Duke has done much to begin supporting early community access and experimentation with new and emerging technologies.  Over the past decade we have developed and strengthened programs that are designed support experimentation at many levels.  The Duke Digital Initiative has supported classroom experimentation with new and emerging technologies and the use of advanced media technology for instruction and training.  Through strategic collaborations with several of our key corporate partners (Apple, Cisco, etc.) OIT has facilitated early access to important hardware and software market innovations. Our Co-Lab, lab computing, and research computing support programs provide the Duke community with opportunities for open and early access to the most current computational tools, techniques and capabilities (3D printing/physical computing, makerspaces, drones, “app’ programming courses/seminars, advanced media labs, virtual/augmented reality, technology consultations, etc.). Notable student-led technology startup companies have taken root in part through the student’s access to new and emerging technologies and to the encouragement and open assistance provided by our program consultants.  We have also provided the university research community with meaningful access to advanced networking technologies through our existing infrastructure investments and through NSF-sponsored activities (i.e.; our research into the deployment of software defined networking infrastructures to enable high-speed and high volume data transmissions between research facilities across the globe, etc.).

3.4.2 Leverage consumer technologies and open source where possible.

The mass market has driven the cost of highly capable electronic devices down to remarkably low levels.  We can create real value for our community while saving time and money by leveraging new consumer technologies where possible rather than developing our own solutions or purchasing low-volume proprietary hardware.  We should also utilize and contribute back to open source technologies where possible.

Implementation will involve refocusing existing staff effort and through activities under 3.4.1 above.

3.4.3 Prepare for the “Internet of Things”.

The “Internet of Things” (IoT) is a development of the Internet in which everyday objects have network connectivity, allowing them to send and receive data. It describes what is today a real trend towards embedding intelligence, sensors, and wireless connectivity into previously inert objects.  Analysts predict that the number of such devices will grow tremendously and we may soon see 30 or more such devices per active social network user as common consumer products begin to incorporate this potentially disruptive technology.  IoT provides great opportunities for innovation and progress across all aspects of campus life.  However, it also raises a number of security-related challenges.  Duke should engage in advanced planning around the expected appearance of many thousands of IoT devices on our networks within the next several years.

CONCLUSION

Duke University, as a world-class research and educational institution with global interests, places ever-increasing demands on the technological environment that is so important to its competitiveness and overall success. As evidenced in this report, the willingness of our community members to come together and participate in a deep and thorough discussion of our IT environment’s strategic strengths, weaknesses, opportunities, and threats is in itself evidence of our community’s excellent capacity to work together toward a set of common goals.  This planning process has allowed participants from across the campus to engage with one another and reach a common understanding of Duke’s technological goals and the strategic actions required to attain those goals. The following are goals and general areas of effort identified in this report:

	Duke Strategic IT Priorities 2017-2021
Help People Use Technology More Effectively ($3.5M annually)	Provide Ubiquitous, Secure Services and Systems ($6.3M annually)	Support Innovation in Research and Education ($3M annually)
Goals:   1.1 Enhance user interface design and user experience across deployed systems   1.2 Improve back-end integration of tools and systems   1.3 Improve web support for university units   1.4 Improve use and support of collaborative tools   1.5 Improve technology in learning and meeting spaces   1.6 Improve our ability to provide accessible versions of digital content	Goals:   2.1 Maintain a modern protected network     2.2 Respond to and anticipate evolving information security threats     2.3 Explore moving key enterprise services to the cloud     2.4 Better integrate our IT purchasing processes with the needs of our users     2.5 Improve ways to request support or consultation between IT units and OIT	Goals:   3.1 Support the evolving computing needs of our researchers     3.2 Improve Duke’s competence in data analytics     3.3 Advance both curricular and extracurricular opportunities for IT education     3.4 Support and promote IT innovation

 

…. = Essential

…. = To meet university strategic planning goals

…. = To meet emerging needs

Achieving this plan involves $2.425M one-time costs and a $12.8M ongoing increase to the base, of which $1.675 is committed within OIT annual funding beginning FY18.  The figure on the following page illustrates the relative ongoing costs across the three interlinked goals of this plan.

 

$1.475M one time

 

***

***$50K annually is already committed by Library