Rensselaer Alumni Magazine - Winter 2013/2014 - (Page 19)
"We need to learn how to harness the force of data, or we will be left
behind as individuals, as academics, and as a nation." J A M E S H E N D L E R
"Here at Rensselaer, we will address the hard problems, which
we are uniquely qualified to address because of our strengths in engineering, science, design, management and entrepreneurship, and
the humanities, arts and social sciences," says President Jackson.
"We will continue to leverage our interdisciplinary approaches to
problem solving and educating students, using the new tools and
technologies of this data-driven, Web-enabled, supercomputerpowered, globally interconnected world."
"We are looking to fundamentally change the life cycle of
scientific inquiry by infusing it and informing it with leading-edge
data techniques," Hendler says. "This holds the potential to touch
every discipline across campus, and in a way that is unique and very
special to Rensselaer."
HANK OSUNA / CONCEPT BY CELIA JOHNSON
SEEKING THE TREE
Rensselaer faculty researchers and students, past and
present, have never been afraid to tackle the biggest,
most complex challenges of the day. This audacity, paired
with the Institute's culture of excellence and interdisciplinarity
in education and research, is a key factor that enables a bold,
strategic endeavor at the scale of the Rensselaer IDEA to take
root and grow.
The new institute builds on key Rensselaer strengths: Faculty in
all five schools are already pursuing data-driven research. Educational innovation is ensuring students start using data and analytics in introductory courses taken within their first few semesters
on campus. Research centers and constellations are developing
critical capabilities and intellectual leadership in the use of data in
science and engineering. And most important, arguably, are the
low walls between engineering and other disciplines on campus.
The situation has evolved, Hendler says, to the point where
barriers to collaboration are no longer internal but external to
Rensselaer. The university has no cultural or procedural impediments to transitioning scientific discoveries into the realm of engineering, where the breakthrough is refined, documented, and
moved forward toward a product or an application.
The challenge, he says, is that very few governmental agen-
cies or private foundations are willing or able to fund this type of
long-term project that spans both basic and applied research across
traditional disciplinary bounds.
This promotes what Hendler calls "the data forest." He characterizes the situation as a forest of slender trees drawing upon
a shared underground aquifer of data. Each tree represents an
individual research project, with the green treetop symbolizing the
data-driven end application. (Such an application could be a piece
of novel software for modeling metal fatigue at the nanoscale, or a
student-created app for wirelessly monitoring home water usage, or
anything in between.)
The challenge is that not all researchers are data scientists or
computer programmers, Hendler says, nor is it reasonable to expect
them to be. This often results in custom-built applications that are
valuable to the intended end users, but which are precariously balanced upon project-specific programming techniques that Hendler
equates to thin, brittle tree trunks.
Even though the underlying data is good, and the program
works, the nature of the solution results in an application that is
expensive to maintain, difficult to validate, and oftentimes impossible to share with others. Eventually, Hendler says, these brittle
data applications diminish and crumble under the weight of their
own complexity and narrow focus.
The Rensselaer IDEA is in the process of reshaping this
forest of delicate saplings into one formidable, spectacular redwood. Here's how the scene changes: Instead of being separate
tiny top-heavy trees, the data-driven applications would be
branches upon the big tree. And the trunk of the tree would be
strong, healthy, and robust enough to support its many branches
indefinitely into the future.
Consider this: An astrophysicist's data, on the surface, looks
considerably different from a network scientist's data, which in turn
looks significantly different from air flow and friction data collected
by a mechanical engineer. "But from a computer's point of view,"
Hendler says, "the data probably doesn't look very different." And
while the tools used to collect this data-a telescope, Twitter, a
wind tunnel-are remarkably dissimilar, researchers often seek to
do very similar things with the information they capture.
This is the opportunity of the Rensselaer IDEA: to create infrastructure, processes, and common tools for manipulating, mining,
and sharing data that can be used by faculty members or students
from any discipline. This common "trunk" will greatly reduce the
RENSSELAER/ WINTER 2013-14 19
Table of Contents for the Digital Edition of Rensselaer Alumni Magazine - Winter 2013/2014
Table of Contents
Snapshot: Rensselaer at Petascale
President's View: A Lesson From Microbes
Feature: The Rensselaer IDEA
Feature: Solving Problems
Feature:Hyerloop, Meet Tubeflight!
Feature: 2013 Alumni Hall of Fame
One Last Thing
Video Clip: Rensselaer at Petascale (AMOS)––Exerpts from President Jackson and CCI Director Chris Carothers remarks made at the Rensselaer at Petascale announcement event (3:21)
Slideshow: GE Girls@Rensselaer (0:54)
Video Clip: Data Forest––Director Jim Hendler discusses the "data forest" at the Rensselaer IDEA annoucement event (2:59)
Vimeo Link: EMPAC Research–– A New Paradigm for Interactive Exploration of Data with Live Coding (23:50)
YouTube Link: Alumni Hall of Fame Induction Ceremony (44:13)
Rensselaer Alumni Magazine - Winter 2013/2014