Duke ITAC - August 25, 2016 Minutes

Duke ITAC - August 25, 2016 Minutes

I.   Announcements

  • Our next meeting will be held in our new alternate location in the Technology Engagement Center in the Telcom building.  The space will promote our student innovation activities, along with research computing efforts.
  • George Ward, who worked with us in OIT, the Libraries, DCRI, and in other capacities around campus over the last 15 years, passed away this weekend.  Those want more information about arrangements for services should contact John Board.
  • Three Zero Day vulnerabilities were disclosed today in iOS.  The three work as a chain to allow attackers to own devices. A patch has been released from Apple and should be applied to your devices as soon as possible as this attack is not theoretical.
  • Both sets of minutes from June 6th and July 28th are approved.

II.   Agenda Items

4:05 – 4:30 – Stadium Technology Updates: Cisco StadiumVision, Bob Johnson, Jack D’Ardenne (15 minute presentation, 10 minute discussion)

What it is:  Efforts to modernize Duke Athletic Facilities have been underway for quite some time.  The updates are designed to enhance the game-day experiences for fans and student-athletes alike.  Features such as new stadium seats, additional seating, a press box, and club suites, new concessions, restrooms and ticketing are all a part of this design.  One of the less known feature upgrades is the technology infrastructure; specifically StadiumVision. 

Why it’s relevant:  StadiumVision by Cisco, is a video and digital content distribution solution to centrally manage and target delivery of customized video, team and sponsor promotions, and relevant event information to any combination of monitors.  Duke plans to use the solution to display menus on TV screens and include a dedicated TV to broadcast the football game.  Bob and Jack will provide an overview of the solution at Duke, along with a live demonstration of its features.

    • Duke Athletics and OIT have been in conversations for quite some time to find a solution to improve on branding and advertising.  The primary goal was to figure out how to release, differentiate, and control high quality content across hundreds of monitors and screens in multiple locations, with a small support staff.  Cisco’s StadiumVision solution is the video backend that provides special effects and advertisements at many major sports venues (Cowboy Stadium, Yankee Stadium, Staples Center, etc.) and optimizes the revenue experience.  Another campus need that could be addressed with this system is a means to facilitate emergency messaging.  One of the areas the system is lacking in is development.  Assistance is needed around the content being displayed, back-end analytics, user interfaces, and the ability to expand the reach beyond campus in the mobile environment; allowing us to move the stadium experience outside of athletic facilities and onto other campus locations. Cisco is also looking for development assistance. Our first football game is scheduled for Saturday September 3rd.
    • What is StadiumVision? – A digital signage/IP TV platform, to target and coordinate advertisements and sponsor promotions with digital and dynamic multimedia displays, dynamic concessions (can be linked to XML feeds to live-update data related to product availability and pricing changes), directional signage (can be used at halftime and at the ends of games for wayfinding of exit routes and restroom facilities), news feeds, all with full display control.  The solution has two main components: StadiumVision Directory and Digital Media Player (DMP)
      • What is StadiumVision Director? – Command center interface that provides the ability to coordinate, manage, and orchestrate video interactivity.  The option is available to pre-program StadiumVision Director and set automated triggers to perform intended actions.  However, most major sporting events have a full-time person driving the content.  The product allows for centralized control of all video endpoints with the ability to segment attached displays into groups independent of location.  Some additional features include a repository for both static and dynamic content, flexible content presentation, automated display of content based on scripting or ad hoc messaging, single button shut off for all TVs in a venue, and the ability to assign role based access for system users.
      • What is DMP? -  All IPTV/content is sent to a digital media player and is fully managed via StadiumVision Director.  Each DMP has a Linux OS, 32GB of internal storage, support PoE (Power over Ethernet) and RS232 control, and come in two models: 4K and 2K.  Duke Athletics primary went with the 4K model.
      • Who uses StadiumVision?  Some examples of professional sports teams and colleges/universities currently using StadiumVision are:
        • MetLife Stadium (New York Giants and Jets)
        • AT&T Stadium (Dallas Cowboys)
        • Staples Center - (Los Angeles Lakers, Clippers, Kings)
        • SAP Center at San Jose (San Jose Sharks)
        • Air Canada Centre (Toronto Maple Leafs, Raptors)
        • Ed and Rae Schollmaier Arena and Amon G. Carter Stadium (Texas Christian University - TCU)
        • Pauley Pavilion (UCLA)
    • Duke Athletics is installing over 350 DMPs throughout their facilities in offices, conference rooms, break rooms, menu boards, outdoors, in suites, the Nike Store in the Scott Family Pavilion, and in the stadium and Kennedy Tower at Wallace Wade.
    • Stadium Vision IPTV - Duke has a 30 channel package with DIRECTV and all channels can be provided to the DMPs via multicast.  The annual fee for channels covers all of Athletics DMPs.  Since this installation requires a one-time subscription fee, the ROI for the project is a little over 5 years. 
    • StadiumVision Demo
    • Other uses for StadiumVision at Duke – Mobile platform, IPTV campus locations, Development of web and mobile apps for controlling DMPs using Cisco’s API, DukeFlyer Integration. 

Questions and Discussion:

Question:  Are the interfaces into this accessible enough for Co-Lab Students?
Answer: Yes
Question:  Can it play the video feed of the game itself?
Answer: Yes.  We have 30 DIRECTV channels with a buffer of 10-15 more open channels to allow for broadcasting of other college games through a sports package. There are also four feeds just for Duke content. 
Question:  Are there possibilities for expansion, HospitalVision for example, where other types of wraps, logos, branding and content are displayed?
Answer:  It can be used as a simple digital signage display.  The installation is setup in zones for campus, hospital, Athletics, with the possibility of creating more granularity by department or building/location.
Question:  Is the current installation set up in zones or are there separate controllers for Cameron vs Wallace Wade?
Answer: It’s all one installation, but zones can be created, along with a more granular option of sub-dividing the zones by location/facility.
Question:  It’s IPTV inside.  Is it overkill to think of a solution like this as a supercharged IPTV or as a supercharged digital signage system? 
Answer: The ability to combine two systems into one is an attractive option.  The problem we’ll see is the potential financial loss on the existing video content endpoints around campus; Roku devices for example.  However, the model shifts in this case with economy of scale. 
Question:   If digital signage is rolled out more across different campus venues, is there a possibility for these displays to be interactive one-on-one?  Could these DMPs and displays be augmented such that in the future a student/faculty member, staff member or visitor could scan a code/device/or other form of identification and the display provide a map and directions to an unknown location on campus? 
Answer: Cisco has made inroads in the professional space and is looking at how to transfer that success to campus environments.  Any ideas like this would be welcomed by Cisco.  There are pros and cons to the existing centralized digital signage solution (Four Winds) and an installation like StadiumVision.  For example, Four Winds is better at touch screen and interactivity, whereas StadiumVision is better at displaying images and video in public places such as West Union and the Bryan Center.  The two however, complement each other, especially now that we have an opportunity for a content repository that can be used by both systems. 
Comment: There may be potential to use something like this in the clinical environment.  However, they may have their own video content provisioning tools.  One of the pioneers in real-time on-screen graphics, tickers, clock-and score, virtual insertions and social media integration providers for live televised sport and entertainment events is within a half mile of Duke; SportsMEDIA Technology[1].  They provide the on-screen yellow line in NFL events for example.  There may be a collaboration opportunity to learn from people who provide production day content. 
Question:  Is there any work towards creating a UI-based function where an end-user would logon to a personal device and it becomes the content screen? 
Answer: Cisco has a mobile platform that allows some integration, but we haven’t vetted the full functionality of the available API and how it can be used to create functions like this. 
Question:  Does the current licensing model allow for members of the Duke community to stream both Duke content and DIRECTV content on their personal devices?
Answer: There are no indications at this time that this feature is available in the current release or in future releases.
Question:  What about non-sporting events (i.e. Commencement) in Wallace Wade?  Alternately, what if there was an emergency accompanied by a severe weather alert?  Is there a button someone could push to show evacuation procedures? 
Answer: The system supports manual triggers for emergency content as external triggers from the National Weather Service or a website. The fundamental capability is there, but the system does not come pre-programmed with this feature.

4:30 – 4:50 – Duke Libraries Data and Visualization Services (DVS), Joel Herndon, Head, Data and Visualization Services (10 minute presentation, 10 minute discussion)
What it is:  Data and Visualization Services provides consulting services and instruction that support data-driven research.  The team offers support in data sources, data management, data visualization, geographic information system, financial data, and statistical software.  Joel will provide an overview of the services DVS provides, with a focus on those that directly relate to research efforts. 
Why it’s relevant:  Data-intensive research is pervading nearly all scholarly fields at duke, including many for which managing, processing, and visualizing large datasets is a relatively new phenomenon.  DVS is a key resource available to researchers at Duke to tap expertise in these areas. 

  • DVS works in three main areas:
    • Consulting on a wide range of data topics
    • Instruction and Workshops – We host about 20-24 workshops each semester on a wide range of data topics for different user communities, and working with Duke classes seeking to integrate data-driven research methodologies and tools. 
    • Computing resources in partnership with OIT and other IT units across campus.  We have a lab in The EDGE with 12 workstations with 32 GB of memory and a Terabyte of shared storage across the machines, along with the latest data visualization, analysis and mapping software.   Since the lab opened in January 2015 has been wildly popular; not only for the computing power on the desktops, but for the large dual displays, roomy working space, and power for laptops. 
    • We have six different support areas
    • Data sources in mapping and GIS
    • Data management is an emerging area.  We’ve been helping people with planning since the National Science Foundation changed requirements for researchers pursuing grants to include a data management plan.
    • Data Cleaning: It’s difficult to do any type of research without some data cleaning.  Where the data is numeric, digital text, or image files, if the data is digital we try to help researchers prepare it for their needs. 
    • Data Analysis: We provide support for a variety of statistical software packages. The most ones most commonly requested are Stata, R, and MATLAB.
    • Data Visualization is a service we started in 2012.  We’ve seen Interest across campus in all disciplines has seen grow exponentially.  Our workshop series has been ramped up to move beyond traditional data visualizations such as charts and graphs to include qualitative data, mapping concepts, and poster presentations.  We have 2 consultants in this area.
    • Demand across the campus is rising for DVS consultations.  Beginning in 2010, we had a little over 800 consults in the academic year.  Now we’re seeing roughly 1,100 questions a semester.  We’ve also seen a rise in the number of 5-10 minute questions, along with an increase 2-2.5 hour questions.  When we take a look at who we’re working with, there is a blend of a lot of new people on campus, looking for data support, have never used data tools before, but know they want to work in data visualization and digital media.  Simultaneously, we’re seeing growth in members of the advanced user community trying to use digital tools and reach more complex results.  We’re seeing a mix of just-in-time demand for course assignment completions, and programmatic initiatives and research grants where a deeper/different level of support is needed.  Both the consulting in the library, the programmatic and event support reinforce each other.  Working with a variety of groups across campus allows us to view the edge of our capabilities and plan better for future endeavors.  For example, we’ve seen an increasing number of people from the Medical Center requiring consulting services around data visualizations.
    • We have worked with a wide range of groups across campus on projects, programming, and different types of events.  Over the last year we’ve worked with institutes, initiatives, and groups such as the Duke Marine Lab, the Duke Energy Initiative, the Digital Humanities, Initiative, the Duke Clinical Research Institute, the at Duke School of Nursing, the Duke Visualization & Interactive Systems Group. 
    • The Workshop Series[2] is used as a way to allow people to get an idea about using digital tools and research methodologies, along with an advertisement about what types of services we provide.    Videos of workshops are placed online using Panopto if live attendance is not an option.  The series covers five of the six different support areas, with data management coming soon. 
      • Workshop Attendance: Multiple identities and affiliations at Duke can sometimes present a challenge when trying to enumerate the different areas and types of people attending the workshops.  The Social Sciences community has historically and currently been the largest number of workshop attendees.  Other groups at Duke rounding out the second and third largest groups registering for the workshops include members of Sanford, Sciences, Duke Libraries, Duke Medicine, Humanities (continues to rise), NSOE, Engineering, and Staff outside the Libraries.  Data seems to be something that in the last ten years, has moved out of being a specialized research tool used in a lab or disciplinary methodology and is now something that a very broad community at Duke is exploring. 
      • Data Management:  We have been working for the last couple of years where we provide advice on data management planning and worked with several faculty members with an interest in storing their research data in Duke’s Institutional Digital Repository.  However, this areas of focus has been explored through pilot projects such as IPHEx[3].    We want to expand this service.  Just this month the Library added a Counsel for Library and Information Resources (CLIR) Post-Doctoral Fellow in collaboration with the Social Science Research Institute (SSRI).  The fellow is working to expand support for curation workflows in education in the Social Sciences for managing data. 
      • Four positions currently posted
        • Senior Research Data Management Consultant (2) – Responsible for expanding training on campus for data curation issues, data curation methodologies, and providing consultations on moving data from a research process into a shared resource for publishing and curation
        • Digital Repository Content Analyst (2) – Responsible for moving content into Duke’s Digital Repository. 
      • We often have a bottleneck of processing all of the information that individuals want to share or preserve and look to these new positions to help us expand on that service to serve a larger research base on campus. Developing Data Curation Services requires more consultations to address individual researchers’ needs. 
      • We want to understand how digital information is produced, described, and consumed, understand the different workflows across Duke, and provide a consultation service that is broadly applicable across campus.  
      • We want to provide curation support by talking with researchers about how to structure a project in a way that makes it easier for both the researcher and the group curating the projects data at the end of the workflow to understand the processes used to create the data and how it can be reused for other projects
      • Increase instruction in Duke Courses is a result of speaking with several faculty members on campus.  The idea is to reach students who are early in the process of learning research methodologies and familiarize them with digital skills to facilitate teamwork on data project and ease of data sharing for publication needs.
      • Workshops are being used to increase programming on campus that sheds light on data curation service needs; whether it’s the mechanics of data sharing or using tools that enable better curation.
Questions and Discussion:
Question:  Are the machines in the DVS Lab being used largely for the high-powered applications they’re intended for? 
Answer: Yes, up until the fall of 2015.  As campus computing resources are reduced, the demand for our lab resources is much greater to serve a diverse user community, including students and researchers.
Question:  How would you characterize the breakdown between students and faculty seeking consulting services?  What are the percentages that represent these groups?
Answer: We see an even split between undergraduate and graduate students (proxies).  We do not see many faculty members coming into the lab.  Some faculty members join us in the library.
Question:  In terms of workshops, one thing that was missing was Python as a tool seems to be getting a lot of traction, but that is not offered as a course. 
Answer: While we have several consultants in DVS that are fluent in Python, we have not yet offered courses in that area.  We are in conversations with Data Carpentry [4] and are debating to offer our training as part of that series or Python.  It is something we could consider pending the audience size.
Question:  DDR is a topic of interest on the clinical side.  Is it being used in production now?
Answer: DDR is the Duke Digital Repository and is not housed in the Data & Visualization Services group.  However, the DDR is currently several different systems:
  • D-Space is the platform we’ve being using to store data sets, but we’re currently migrating to Fedora[5] as our storage infrastructure and have loaded Duke Libraries data sets into it, but have not begun loading faculty research data.  Migration of faculty research data will begin soon so researchers won’t be concerned about platform storage infrastructure.

Question:  You mentioned services like data cleaning.  Are your activities in this domain mainly consultative or are they a service bureau kind of function where a researcher can bring in data and have it cleaned, scrubbed and anonymized? 
Answer: Our services are consultative.  However, our colleagues at SSRI anonymize data.  We are currently discussing different tiers of curation.  Not all data projects require, nor are all researchers seeking a high level of curation.  But we know we aren’t equipped to scale all projects at a deep level of curation.  There is a wide range of interest on campus with everything from basic citation (a very minimal about of meta data), a more moderate set about structuring information in a way for secondary data use, and a rich level of data cleaning both for machine readable data and human use. The two Research Data Management Consultant positions were created to educate people who are immersing themselves in data management plans and are getting ready to start research projects to think about the best practices, formats and structures along the entire research project.  While the two Digital Repository Content Analyst positions are– analysts to help put content into the repository and performing the QA pieces to help ensure those best standards are being followed, but also to advise those who are depositing information into the repository that might be disruptive with a self-service support model allowing faculty and researchers to flow the data through.
Question:  With all the groups around, yours being a big one, SSRI, IID activities, we have lots of great activities, but undoubtedly there are still gaps.  From your vantage point, what would you say if we had two or three things that are huge gaps for Duke as a whole?   What are they in this space? 
Answer: The first thing we’re thinking about it raising awareness of what the curation issues are and what the grant and publishing landscape looks like.  We were concerned about granting agencies four years ago when we were talking about data management planning.  Secondly, NSF requiring data management planning rules, and then the office of science and technology policy requirement for most government grants that data sharing would be a part of them.  In the first few years, requirements enforcement was uneven with many of the agencies.  However, there seems to be a shift in the past year to a more rigorous model.  We’re also seeing that a lot more publishing in journals that require data management plans than there are grants at Duke.  We suspect that is going to be the larger portion of data curation.  We want to expand support model to address these needs, preferably through instruction, to virtually eliminating last minute publishing snafus by providing advising services up front. The other areas we’ve talked about in the past is how much curricular systems faculty has asked our labs to provide students with the education they need about how to manage data and data planning.  We don’t always get the advanced word when curriculum is moving or adapting in an area of need.  For example, an entire section has shown up during a week period asking to learn about something, but there hasn’t been any communication between the professor and our group.
Question:  How is your staffing model set up for consultations on data visualization services? Do you have the capacity to do online consultations since we have a lot of students and faculty that are remote? 
Answer: We have the capacity to do that, but don’t have a formal service model that lists certain hours.  We do, however, have plenty of facilities where we can leverage remote technologies such as Google Hangout and Skype.  We utilize streaming technology as well for researchers at the Marine Lab, DKU, and for students that are away from campus. As we see more demand, we’ll work to provide a means to provide the service. 
Question:  Is on-campus consulting done generally on a walk-in basis? 
Answer: Yes, but if it’s a faculty member and there is a high demand for an on-site consultation, that is an option.  The lab was designed to enable us to have the right tools to provide the service.  However, the increase in Virtual Machines (VMs) provided by OIT, along with Docker Terminals, afford us the option of providing support in the absence of the lab.  Contact our office at askdata@duke.edu for more information.
Question:  Is there a way to share presenter information on the ITAC website? 
Answer: We can discuss and figure out a way to do it.

4:50- 5:10 – Duke Innovation Challenge: Tackling the Zika Virus, Jackie Xu, Francia Fang, Howard Li, Michael Li (10 minute presentation, 10 minute discussion)
What it is:  SEAD (The Social Entrepreneurship Accelerator at Duke),  the Center for Advancement of Social Entrepreneurship, the Duke Innovation and Entrepreneurship Initiative and the Duke Global Health Institute The Zika Innovation Challenge, held in March of this year aimed to contribute tangible ideas and actionable solutions that might mitigate the challenges related to the Zika epidemic in the Americas, and global health security at large.  Teams of Duke University students were challenged to address three major focus areas:

    1. Boost Tracking & Surveillance Activities
    2. Awareness Among & Protection of Pregnant Women
    3. Integrated Vector (i.e. Mosquito) Control
Why it’s relevant:  The winning team, composed entirely of freshman, proposed using social media platforms to alert travelers to the disease and its risks, especially during and after the 2016 Summer Olympic Games in Rio De Janeiro.  Their plan was to warn travelers who make posts from Brazil, direct them to get tested and take a survey that researchers could use to identify risk factors and alert local public health agencies.  The team presented their idea to the U.S. Agency for International Development’s Center for Accelerating Impact and Innovation, and wants to establish a working model of their proposed solution in the future.
  • Introduction: We are all sophomores who met while working through GlobeMed[6], where we first heard about the Zika Challenge.  The challenge was started by the USAID (United States Agency for International Development) [7]and was designed to get ideas for how to innovate and solve some issues around Zika.    The challenge has held over the course of a weekend were we attended workshops on our focus: tracking and surveillance of the Zika virus.    After putting our head together, we decided to use social media to track and surveil the virus.  After presenting to a panel of judges from all corners of Duke, we won first place.  From there, we were asked to present at the Alliances and Incentives in the Era of Outbreaks Symposium, where we were able to network with a representative from the Bill & Melinda Gates Foundation and received valuable feedback regarding how to get funding for our project.  Finally, we became of another challenge hosted by the USAID: Combatting Zika Grand Challenge.   The challenge was open to people of all ages and professions.  After submitting a revised version of our proposal from the Duke Innovation Challenge, we were selected to present at our idea at the USAID headquarters in Washington, D.C. this summer.  One of the highlights of the trip was the opportunity to meet with Dr. Anthony Fauci[8], the director of the NIAID (National Institute of Allergy and Infectious Diseases).[9] Not only is he a lead expert on infectious diseases, but has worked on numerous presidential advisory committees and was one of the principal architects of the President’s Emergency Plan for AIDS Relief (PEPFAR) program and was able to offer us valuable feedback on our presentation.  We presented in in front of panel of judges consisting of members of the USAID, the Department of Defense, Latin American Government officials, along with industry executives.  While we aren’t looking to USAID for further support, but it is still a project we want to pursue. 
  • The Project – Zika has quickly become a major health concern.  Currently there are two issues with tracking and surveilling the virus:  First, public health officials have difficulty collecting data on the rates of transmission. Secondly, the public is still relatively unaware of the dangers of Zika.  We’ve found that traditional methods for data collection are outdated.  Currently, if someone has Zika, and goes to the doctor, the doctor reports it to public health agencies.  Additionally, only about 25% of people show symptoms at all and the majority of people who are infected don’t know it. This is especially a problem for pregnant women as Zika has been linked to babies born with Microcephaly and other birth defects.  
  • Impact Gap Canvas
    • Challenges: Zika virus carriers lack symptoms, transmission methods are unknown, the virus has been linked to microcephaly, and very few people receive diagnostic blood tests
    • Impact Gaps: There is a lack of information on the spread of Zika in Brazil and worldwide, there is a lack of prevention measures, and there is a lack of knowledge and diagnostic tests
    • Current Solutions Mapping: Doctors manually report virus instances is inefficient and blood tests for Zika are rarely conducted.
    • We needed something that provided real-time disease transmission data to public health officials while simultaneously providing informational resources to the public about the dangers of Zika. 
    • Proposed Solution Package
    • Our target audience was the 2016 Rio Olympics because Brazil is a well-known Zika hotspot and the Olympics attracted 500,000 visitors to the area from all over the world, allowing Zika to potentially spread worldwide. 
    • Our idea was to use location monitoring services in social media platforms such as Twitter, Facebook and Instagram, to monitor when people travel to Zika hotspots.  For example, if someone traveled to Rio for the Olympics, they would likely post pictures and statuses with their location.  This same information could be collected for public health use.  Once the social media site detects the location posts in a Zika hotspot area, a quick survey could be sent to the person with questions to determine the risk at which a traveler has been exposed to the virus.   (Have you been bitten by mosquitoes recently?  (Yes) Have you developed rashes, fevers, or joint pain? Etc.)  At the end of the survey or before the traveler returns home, if it is determine that they may have contracted the Zika virus, the social media site will send the traveler an important message including the dangers of Zika, and additional resources to help them get treatment.  Simultaneously, the survey data and the location tracking data can be accumulated and sent to public health officials at the World Health Organization to help them monitor the spread of the disease.   
    • Who benefits?  There are potentially two main benefactors to a system like this:
      • Individual users being able to communicate information on Zika with others
      • Health officials collecting this data allows them to identify high risk individuals for Zika, and also identify the locations where these high risk individuals are.  This will allow larger organizations like the World Health Organization or the Centers for Disease Control and Prevention to work closely with local health officials to help take needed preventive actions to keep people better informed.
      • Why our solution?  Our objective is to track and stifle worldwide spread of Zika to be more proactive and preventive and contribute to the worldwide understanding to disease transmission.  One of the strong points of the solution is the use of existing technology.  Social media is a pre-established platform that has created a large data collection base and is a very personal method of communication.  One of the more concerning points of a system like this is the use of data location data.  There are, however some existing precedents on social media that address this issue; Facebook Safety Check, for example.  Safety Check allows users to inform their friends and family networks that they’re safe after a disaster or emergency situation occurs.   When something like that happens, Facebook sends users in the area a notification asking them if they’re okay and whether they want to inform people around them of their status.
      • Advantages and applicability – This is a crucial opportunity to sample an international community.  Traditionally public health data is gathered by deploying people to locations and using broken systems of data collection that don’t have much infrastructure support.  Sampling international events provides the opportunity to collect data on a large sample easily and track disease outbreaks as attendees return to their home countries.   Since we plan to use existing technology, the implementation process would be simple and easily scalable to any type of gatherings. (World Cup, Ball Drop on New Year’s Eve, etc.)  We foresee technology revolutionizing the field of public health, driving innovative solutions across the industry. 
      • Simple implementation process that is easily scalable to any type of gathering.  Technology will have a 

Questions and Discussion
Question:   You’re proposing a platform that can be expanded beyond Zika to any kind of health outbreak.  However, the key to success for you is to hook into the social networks.  Have you had any success with generating interest from these entities for this kind of endeavor?    
Answer: While we have had some contact with them via alumni connections, there wasn’t a lot of interest in pursuing a partnership with us on this project.  Our project’s strength is also it’s weakness; utilizing existing technologies is only useful when the technology is available for us to use it in the way we’re proposing. 
Question:  Don’t these technologies expose APIs that can be accessed and are they available for public use? 
Answer: Yes, however, people have to opt-in/subscribe to use it. 
Question:  What about very poor communities that are more liable to have higher cases of this disease where no network (ISP, wireless, cell phone coverage) is available?  Can we assume that everyone has a cell phone or is on Facebook?   Another potential concern is there are select groups that are traveling to large international events (World Cup, Olympics, etc)
Answer: We have looked into the usage of social media in more remote areas of the world and it is still predominantly used.  Other applications where this solution can be deployed are WhatsApp[10] or WeChat[11] in many places in the world. 
Comment: Internet access in Africa and China goes mostly through mobile devices. 
Question:  Is there a certain target demographic that might be missed? Are grandmothers as well covered as grandchildren?   
Answer: Similar to the US, grandmothers are not a member of the target sample.
Question:  What kind of feedback did you get from the NIH in terms of privacy information where people are opting into providing their travel information and then an after the fact communication related to an outbreak and at-risk levels is disseminated?  Were there a lot of questions about that?
Answer: We focused more on how the information could be collected than the privacy aspects.  We focused more on the screening questions for gathering transmission/risk data, rather than user privacy.  We got a chance to look into the privacy policies of the major social media players (Facebook, Twitter, Instagram) and they are lenient.  We found that there are workarounds that we could implement into the solution: De-identification of personal information, working with cell phone companies to anonymize cell phone data, or using a voluntary user agreement to disclose survey result for the sake of public health. 
Comment: It would be interesting to see if the government would support this project because of the stigma associated with it.

5:10 – 5:30 – IT Strategic Plan Update, Julian Lombardi, John Board (10 minute presentation, 10 minute discussion)
What it is:   Under the direction of the Provost, Duke is undertaking its first institution-wide strategic planning effort in about ten years.  Strategic planning around the evolution of IT systems and services, both central and distributed, will help inform the larger university planning process.
Why it’s relevant:  Following the model successfully used for the last plan, seven IT strategic planning groups were formed to address specific aspects of the IT evolution on campus.  Julian and John will discuss the process to collate and distill the good ideas from the various planning groups into a coherent higher level IT plan, along with next steps.

  • The plan is comprehensive for IT at Duke, not just OIT.
  • The plan was developed by collecting stakeholder input from seven working groups.  Those groups developed a set of reports with 43 specific recommendations.  The reports were passed on to the IT Strategic Plan Steering Committee who then consolidated the recommendations into a smaller sets of more actionable items in the form of three top level goals:
    • Help people use technology more effectively – Technology has to be easy to use and has to work on any device.
    • Provide ubiquitous, secure services and systems – Duke is a leader in supporting innovation and embracing appropriate uses of emerging technologies and collected data. 
    • Position Duke as a leader in supporting innovation and embracing appropriate uses of emerging technologies and collected data.
    • There are 15 sub-level goals across the 3 top level goals that will be used as action items for implementation. Budgets, timelines and implementation plans will be created.  Those implementation activities will be informed by the original working group reports.  Codes were assigned to the 43 recommendations to ensure they were included in the 15 sub-level goals. 
  • Sub-level activities for Goal 1
    • Enhance user interface design and user experience across deployed systems by incorporating modern principles of user-centered design via a campus-wide coordinated approach was a common theme across all of the working groups. 
    • Improve back-end integration of tools and systems so systems can talk to each other and users can move from system to system fluently. 
    • Duke has an uneven level of capacity in maintaining modern, accurate, secure websites, where some are much less well funded than others.  
    • Improving the use and support for collaborative tools
    • Improving classrooms
    • An item added by the steering committee addresses the issue of notion of accessible content for the media being created, media currently being used, and provisioning the resources so that new media can be created.
    • Sub-level activities for Goal 2
    • Maintain a modern protected network is a common theme in many working groups as well.
    • Support evolving computing needs of researchers
    • Respond to and anticipate evolving information security threats
    • Explore cost and benefits of moving key enterprise services to the cloud
    • Better integrate IT purchasing processes with our end users
    • Provide more effective ways to provide support or consultation between IT units (IT-to-IT support)
    • Sub-level activities for Goal 3
      • Position Duke as a leader in innovation.  We are already regarded by our peers as leading the way in this area with projects like Co-Lab and the innovative networking we’re doing.  The goal is to promote a culture of IT innovation across the Duke community.
      • We have lagged behind some of our peers in our competence in exploring data analytics for both educational and operational purposes. 
      • Advance curricular and extracurricular opportunities for general IT education.
      • Next steps – We’re placing dollar values on each of the sub-level goals and will bring that back to the Steering Committee until the iterations lead to having a document that contributes to the larger University Strategic Plan and serves for our own purposes as a blueprint for some activities in OIT and other IT units across campus 

Questions and Discussion
Question:   How do you assign monetary values to these? 
Answer: It’s a best guess at what a reasonable number is to achieve the goals whether via tools or additional man-power to expand the capacity of a service. 
Question:  What is the timeline? 
Answer: The plans timeline is 5-10 years.  However, IT generally heirs on a shorter timeframe for anything that isn’t completely speculative.  We have plausible planning horizons out to 5 years for things we know we need to know.  IT also comes with a set of unknowns that are not in the plan.  We need people to be thinking about and exploring emerging technologies such as the Internet of Things (IoT).
Question:  How much dialog happens between the strategic plans of separate entities?  For example, OIT doesn’t have anything to do with classrooms.
Answer: The IT support people in the schools do and are members of the groups contributing in the development of this plan. 
Comment: While the earlier presentation on digital displays and their amazing capabilities was intriguing, there is a spot with holes in the wall where the digital board used to be in the Languages building.   It was taken down and has not been replaced.  There is a discrepancy between schools and the alienation between technology implementations and the rollout to the schools with the varying support models across the IT groups.  How is the funding for the rollout of these technologies across the schools, and the varying IT units that serve them determined?  Is there a way to apply a standard evenly across the university?
Comment:  There is good and bad in our decentralized IT environment.  The challenge can be keeping the best of both worlds—the benefits of centralized economies of scale and scope and consistency, along with the flexibility and local “touch” of decentralized groups.  The goal is to get to the point where there is a common baseline of technological availability that can be relied upon, even in our decentralized environment.  Goal 3 may need a 4th sublevel activity to work for establishing an appropriate level of baseline services campus wide.  We can’t be successful as innovators if access to basic technologies is too uneven at the school and department level.


[1] http://www.smt.com/ 

[2] http://library.duke.edu/data/news

[3] https://iphex.pratt.duke.edu/

[4] http://www.datacarpentry.org/about/

[5] http://fedorarepository.org/

[6] https://duke.collegiatelink.net/organization/globemed

[7] https://www.usaid.gov/

[8] https://www.niaid.nih.gov/about/directors/biography/pages/biography.aspx

[9] https://www.niaid.nih.gov/Pages/default.aspx

[10] https://www.whatsapp.com/

[11] https://web.wechat.com/