I watched our talented students develop a user-friendly interface to load large data sets into their ParaView application so that we could visualize geophysical data via a high definition 3D display. This is representative for a heavily focused STEM university where we have a number of research applications from geoscience, environmental and industrial engineering sectors that present us with large diverse data sets. We also have campus leadership who feel we should create a strategy to keep pace with the Big Data Science opportunities. So to stay in the mainstream, we created courses that responded to the Big Data buzz primarily based on the various Hadoop solutions. Fortunately, we have a strong relationship with corporate partners because of our SAP based ERP program, so when we realized that SAP HANA was the leading contender for their Big Data solutions, I started exploring how we could matchup our teaching and research efforts with these corporate partners. All that to say we had potential galore, but how does our university best respond?
It was not difficult to confirm that our corporate partners were intrigued by our interest in HANA, but working out a mutually beneficial relationship was more complicated. We could see potential success coming from our ability to interpret large data sets to identify anomalies for research value, but that tended to be data sets over a large time span. Big Data is about the ability to quickly visualize and analyze large and varied data sets. We have watched the petroleum industry define their version of this, but what about manufacturing, heavy equipment maintenance, environmental scanning or transportation.
We have the intellectual expertise to solve these problems, but not with the Big Data tool sets being explored by our corporate partners. So how do we bridge that gap?
Research in Higher Education has been going through a dramatic shift based on the change of funding sources to the ability to staff research teams with graduate students. The shift to the corporate relationship model requires a responsive solution driven by market trending technology. The previous government funded model was more about having the luxury to discover the best technology. The corporate collaboration model has grown out of relationships typically connected to recruiting of graduates. I believe university research also needs to open new doors with corporate partners to uncover research collaboration opportunities. We must at least stay on track with the research and development that the private sector is involved with if we hope to stimulate the new research funding that we need.
With this HANA example we try to insert a way for our researchers to respond to corporate collaboration opportunities that requires proficiency in their tool of choice. Our researchers don’t have the ability to procure HANA, let alone find grad student expertise to run it. Here is where I see the higher education model changing. Research support leveraging the breadth of
Information Technology Services needs to bridge the gap. In the HANA example it requires a commitment of staff DBAs and Analytics professionals who must become proficient in how to utilize this specific version of a Big Data management tool. That expertise allows IT to translate the needs of the customer and assist the researcher in responding to the opportunity.
Involving the IT organization in support of technology used in research may seem very straightforward to the corporate mindset, but this has not exactly been the case in higher education research. Every model imaginable exists but the norm falls mostly with research centers providing for their own needs.
IT historically took care of the network connection and occasionally got involved computational support. The more traditional High Performance Computing (HPC) campus support model would tend to have a more direct involvement with IT, but it too is kept at a distance. However, this HPC support model has provided the template for moving researchers to seek more direct assistance from IT experts. In fact I see IT getting more involved as the facilitator for connecting researchers with national computing resources such as XSEDE.
Big Data is about the ability to quickly visualize and analyze large and varied data sets
Is there a trend here? Higher Education is experiencing a number of troubling trends relating to their business model. Their products are not performing well. The student’s return on investment is being questioned and the Ivory Tower approach may be at blame. Where does a CIO fit into this situation? University Information Technology has mostly been a service organization working behind the scenes to help faculty, staff and students keep pace with technology. IT may rise to visibility during implementations of enterprise software boasting solutions that will reverse these trends; however, IT is also commonly crucified for the failure of such efforts. In many institutions the established academic leadership fears IT, but it is a good fear. It is a fear based on how important technology has become in the business of conveying knowledge.
Many university CIOs will argue that they should be given a greater leadership role to have a chance to solve some of these challenges. But the reality is that the Ivory Tower is also appealing to the university CIO. Change will not come from university leadership, CIOs will need to show them the way and that will mean leaving the comfort of the Tower.