Cyber-enabled
Discovery and Innovation
(CDI)
Objective:
Enhance American competitiveness by enabling
innovation through the use of computational thinking
Cyber-Enabled Discovery and
Innovation
Multi-disciplinary research seeking contributions
to more than one area of science or engineering,
by innovation in, or innovative use of
computational thinking
Computational thinking refers to computational…
…Concepts
 …Methods
 …Models
 …Algorithms
 …Tools

CDI is Unique within NSF
five-year initiative
all directorates, programmatic offices involved
to create revolutionary science and
engineering research outcomes
made possible by innovations and advances
in computational thinking
emphasis on bold, multidisciplinary activities
radical, paradigm-changing science and
engineering outcomes through computational
thinking
CDI Philosophy
“Business as usual” need not apply

“Projects that make straightforward use of existing
computational concepts, methods, models,
algorithms and tools to significantly advance only
one discipline should be submitted to an appropriate
program in that field instead of to CDI.”
No place for incremental research
Untraditional approaches and
collaborations welcome
NSF Review Criteria
Intellectual Merit
Broader Impacts
New on Transformative Research: to what
extent does the proposed activity suggest
and explore creative, original, or potentially
transformative concepts?
Additional CDI Review Criteria
The proposal should define a bold multidisciplinary research agenda
that, through computational thinking, promises paradigm-shifting
outcomes in more than one field of science and engineering.
The proposal should provide a clear and compelling rationale that
describes how innovations in, and/or innovative use of, computational
thinking will lead to the desired project outcomes.
The proposal should draw on productive intellectual partnerships that
capitalize upon knowledge and expertise synergies in multiple fields or
sub-fields in science or engineering and/or in multiple types of
organizations.
potential for extraordinary outcomes, such as,
 revolutionizing entire disciplines,
 creating entirely new fields, or
 disrupting accepted theories and perspectives
… as a result of taking a fresh, multi-disciplinary approach.
Special emphasis will be placed on proposals that promise
to enhance competitiveness, innovation, or safety and
security in the United States.
Long-term Funding for Cyber-enabled
Discovery and Innovation
All NSF directorates are participating in this activity
(subject to budget approval); estimated $750M
investment in 5 years:
Request
FY 2008
FY
2009
$52M
$100M
($26M in
the
solicitation)
FY 2010
FY2011
FY 2012
$150M
$200M
$250M
Three CDI Themes
CDI seeks transformative research in the following general
themes, via innovations in, and/or innovative use of,
computational thinking:
From Data to Knowledge: enhancing human
cognition and generating new knowledge from a wealth
of heterogeneous digital data;
Understanding Complexity in Natural, Built, and
Social Systems: deriving fundamental insights on
systems comprising multiple interacting elements; and
Building Virtual Organizations: enhancing discovery
and innovation by bringing people and resources
together across institutional, geographical and cultural
boundaries.
From Data to Knowledge
Knowledge extraction, noise, statistics
Modeling, data assimilation, inverse
problems
Validation; model/cyber/domain
feedbacks
Algorithms for analysis of large data
sets, dimension reduction
Visualization, pattern recognition
Understanding Complexity in
Natural, Built, and Social Systems
Identifying general principles and laws that
characterize complexity and capture the
essence of complex systems
Attaining the breakthroughs, to overcome these
challenges, requires transformative ideas in
the following areas:
Simulation and Computational Experiments
Methods, Algorithms, and Tools
Nonlinear couplings across multiple scales
Virtual Organizations (VOs)
Design, development, and assessment of VOs
Bringing domain needs together with algorithm
development, systems operations,
organizational studies, social computing, and
interactive design
Flexible boundaries, memberships, and
lifecycles, tailored to particular research
problems, users and learner needs or tasks of
any community, providing opportunities for:
Remote access
Collaboration
Education and training
Types of Projects
CDI defines research modalities
Project size not measured by $$
Projects classified by magnitude of
effort
Three types are defined: Types I, II,
and III
Type III, center-scale efforts, will not be
supported in the first year of CDI
Type I Projects
focused aims that tackle
discrete, high-risk problems
that, once resolved, may
enable transformative
breakthroughs in multiple
fields of science or
engineering through
computational thinking
research and education
efforts roughly comparable to
that of up to two
investigators with summer
support, two graduate
students, and their research
needs (e.g., materials,
supplies, travel), for a
duration of three years
Type II projects
multiple major aims that tackle
complementary facets of complex
solutions for advancing multiple fields
of science and engineering through
computational thinking.
several intellectual leaders,
multidisciplinary teams
significant education component
likely to be distributed collaborative projects with more
extensive project coordination needs
greater effort than in Type I, and, for example, roughly
comparable to that of up to three investigators with
summer support, three graduate students, one or two
other senior personnel (post-doctoral researchers,
staff), and their research needs (e.g., materials,
supplies, travel), for a duration of four years
Type III Projects
collaborative research, potentially distributed across several
institutions
may involve center-type activities, demanding substantial
coordination efforts
greater effort than in Type II in terms of scope and in the
order of magnitude of expected outcomes
Type III projects will not be supported in FY08, but in the
future years, subject to the availability of funds
Broadening Participation
diversity of sciences
and engineering,
academic departments
underrepresented
minorities in STEM
collaborations with
industry in order to
match


scientific insights with
technical insights
International Collaborations
involve true intellectual partnership in which
successful outcomes depend on the unique
contributions of all partners, U.S. and
foreign
engage junior researchers and students in
the collaboration, taking advantage of cyber
environments to prepare a globally-engaged
workforce
in conducting research in all of the major
components of the CDI
create more systematic knowledge about
the intertwined social and technical issues of
effective VOs, changing both the practice
and the outcomes of science and
engineering research and education.
NSF awards are, in
principle, limited to
support of the U.S.
side of an
international
collaboration. In
almost all cases,
international partners
should obtain their
own funding for
participation.
Examples
Cyber-enabled discovery and innovation in any field
of science or engineering is appropriate for the CDI
program.
Examples illustrate desired outcomes of potential
successful CDI projects.
Note: included for purposes of illustration only; it
is neither exhaustive, nor indicative of preference
regarding research areas.
The listed examples represent contributions in one
or more CDI themes, via multidisciplinary
approaches that hinge on innovations in, or
innovative use of, computational concepts,
methods, models, algorithms, and tools.
http://www.nsf.gov/crssprgm/cdi/
Key Dates:
Letters of Intent (required) due:
Nov 30, 07
Preliminary Proposals due:
Jan 8, 08
Full proposals due:
April 29, 08
 Full proposals by invitation only!
Awards: no later than October 2008
More Information on CDI:
Contact members of CDIIT.

Contact the CDI Co-chairs Sirin Tekinay
(CISE), Tom Russell (MPS), Eduardo
Misawa(ENG) or members of the team
listed in the solicitation
cdi@nsf.gov ; (703) 292-8080
http://www.nsf.gov/crssprgm/cdi/
Questions? Comments?
Example:
Systems with many interacting parts on multiple scales require major
advances in modeling to limit the interactions to the essential ones, in
algorithms to solve the models efficiently and accurately, in software
implementation of the algorithms, and in analysis of the massive generated
data if the challenges of length and time scales are to be overcome.
Examples. Researchers have visualized the changing atomic structure of a
simple, plant-infecting virus and revealed key physical properties by
calculating how each of the virus' one million atoms interacted with each
other every femtosecond. A few additional examples of other domains
with analogous challenges: in chemistry, in analyses of molecular structure
and the dynamics of excited states; in astronomy, in modeling of galactic
interactions; and condensed-matter physics and materials engineering, in
custom design and synthesis of tailor-made molecules for new materials
such as fibers, coatings, ceramics, and electronic materials.
Simulation results may in turn suggest that the theory underlying a model
requires revision; the Einstein equations of general relativity may be an
example. The challenges are compounded by stochastic behavior, which
may arise because of the fundamental nature of a system at some scale
(for example, a quantum mechanical model), data uncertainties or noise,
or incorporation of effective small-scale phenomena into large-scale
models.
Themes: Complexity, Data to Knowledge.
Domains: all fields of science and engineering.
Example
The semiconductors and magnetic memory materials from
which today’s computers are built, and the nanoscale physical
processes used to fabricate them, are the results of
fundamental research in the physical sciences. This research
has fueled Moore’s law. Moreover, new materials, like photonic
bandgap materials and higher density memories, hold great
promise. Even more striking are the completely new
approaches to computer design on the horizon, for example,
quantum computing, molecular computing, and spintronics.
These offer the possibility of revolutionary advances, and
constitute the best hope for continuing, or accelerating, Moore’s
law of hardware into the future. Novel large-scale simulations
based on advances in computational models, methods, and
algorithms play a key role in implementing these approaches,
through fundamental understanding of the nanoscale and of the
emergence of macroscopic properties. This creates a virtuous
circle: cyber-enabled discovery that, in turn, enables cyber.
Themes: Complexity, Data to Knowledge.
Domains: physical sciences, engineering, computer science,
mathematical sciences.
Example
Physical, electrical and cyber infrastructures, such as drinking
water and wastewater treatment facilities; electrical energy
generation, transmission, and distribution systems; chemical
production and distribution systems; communications networks;
transportation systems; agriculture and food production; and
public health networks, are critical to the nation's welfare,
security, and ability to compete in a global economy. Research
to date has separately considered the issues of resiliency,
sustainability, and interdependence. Complexity issues include
cyber-enabled methodologies to analyze and forecast how
infrastructures grow, self-organize, interact, renew, and operate
as interdependent resilient and sustainable systems.
Interdisciplinary, geographically diverse, virtually connected,
nonlinear dynamic networks that predict and control changes
across multiple infrastructures, length and time scales, with
fidelity and the ability to handle huge volumes of data could
involve a large number of disciplines and organizations.
CDI themes: Data to Knowledge, Complexity, Virtual
Organizations.
Domains: engineering, computer science, social sciences,
physical sciences, biological sciences.
Example
Manufacturing in the U.S. is affected by globalization, environmental
and safety restrictions, and competition from an improved foreign
scientific workforce. Some recent developments of interest to
researchers in engineering include just-in-time production, assembly
and/or delivery, shortened product life cycles, and demand for zerotolerance operational incidents. Simultaneously, analogies from the life
sciences are motivating the design of self-assembling and self-repairing
materials. This could lead to the design and manufacture of new
materials and devices, such as artificial skin and self-optimizing fuel
cells. Interaction between researchers in these and other potentially
relevant fields would benefit from novel mathematical and
computational thinking, from complexity analysis, and from
geographically disparate virtual organizations. Combining research in
multi-scale dynamic modeling and simulation for synthesis, design,
prediction and control; large scale optimization; product allocation;
data interoperability; sensor networks; organic and inorganic
chemistry; materials synthesis; and device fabrication are relevant CDI
topics. Research and education projects in some of these areas are
ideally suited for industry/academic collaborations, which might be, but
are not limited to, GOALI projects
(http://www.nsf.gov/pubs/2007/nsf07522/nsf07522.htm).
CDI themes: Complexity, Virtual Organizations.
Domains: engineering, materials science, mathematics, chemistry,
biological sciences, computer science, education.
Example
Living systems function through the encoding, exchange, and
processing of information. The discovery of genetic code and
the ability to capture it in digital form has transformed biology
by catalyzing the creation of databases and applications for
understanding the meaning of genetic code, to compare it and
to predict its function. New research seeking similar
understanding of the communication flowing at other systemic
levels such as chemical pathways, cell signaling, mate selection,
or ecosystem services feedback poses a challenge to
information science to develop more advanced cyber tools for
digitally representing and manipulating the increasingly complex
data structures found in natural and social systems.
Themes: Complexity, Data to Knowledge.
Domains: information science, biological sciences, social
sciences, physical sciences, mathematical sciences.
Example
Theoretical foundations offering tools for understanding, modeling, and
analysis of large-scale, complex, heterogeneous networks of signaling,
signal processing, computing, decision making, communicating,
sensing, controlling nodes with multi-scale interactions need to be
developed. Network science, drawing from economic theory, multiscale analysis, and network information theory, is currently in its
infancy. The Internet, with its billions of interfaces, and mobile,
wireless devices, spanning from personal area networks to satellite
communications, cuts across man-made, social, and natural systems.
Specifically, communication networks, wired and wireless, span the
globe and have become an indispensable tool for modern society,
including science and engineering. New models and analysis tools are
needed to understand spatial and temporal behavior of interactions in
the electromagnetic medium, or in the routing and resource allocation
level. Another area is biological networks, whose understanding
remains rudimentary. New, realistic models of signal transduction
pathways, incorporating interactions with other pathways and behavior
under prolonged stimulus or lack thereof, are needed. Other topics
involving complex coupled networks include communication systems,
the human brain, and social networks. All of these cases call for better
understanding of network structure, function, and evolution.
This example spans all three CDI themes: massive sets of network data
should produce knowledge of patterns across many temporal and
spatial scales; networks, man-made, social, or natural, embodiments of
Example
Develop techniques to forecast critical events in geophysics and predict their impact on society. Central is
the ability to adaptively configure the resolution of numerical models and real-time observing networks; to
zoom in and follow important dynamic features (ocean eddies, earthquakes, volcanic eruptions, landslides,
storms, flash floods, hurricanes, algal blooms, etc.); and to predict their impact on human society,
infrastructure, and ecosystem services. Capabilities such as the tracking of hurricanes necessarily involve
uncertainty, due to the intrinsic nature of the dynamics, limited understanding of features such as the
coupling between the ocean and the atmosphere, and constraints on resolution of practical computations;
quantifying and managing the uncertainty is of critical importance. Themes: Complexity, Data to
Knowledge. Domains: geosciences, ecology, mathematical sciences, social sciences, engineering.
Model, simulate, analyze, and validate complex systems with large data sets. Extraction of significant
features and patterns from high-dimensional data, which can be noisy, is crucial in a great variety of
settings. Examples include the Earth system (geosciences), gravitational waves (physics), galaxy formation
(astronomy), highly complex dynamical systems simulation, health monitoring, prediction, design and
control (engineering), communication and network control and optimization (information technology),
human and social behavior simulation (social sciences), disaster response simulation and anti-terrorism
preparation (homeland defense), design of smart systems for mitigation of exogenous threats using
autonomic response (homeland security), predictive understanding of ecological and evolutionary
processes at multiple scales (biological sciences), software development (information technology), and risk
analysis. A key issue for some systems is understanding whether they will enter a fundamentally different
mode of behavior when an input crosses a tipping point; examples include the Earth’s climate (due to
atmospheric carbon dioxide) and the U.S. economy (due to the federal funds interest rate).
Themes: Data to Knowledge, Complexity.
Domains: all fields of science and engineering.
Example
As hypotheses in the social, behavioral, and economic sciences
have become more sophisticated, so have basic data needs.
Merging biomedical data with survey and administrative data is
a relatively untested area, but it is becoming more crucial for
understanding hypotheses emerging from behavioral economics
and other fields. Understanding human/environmental
interactions requires the merging of data across multiple scales,
such as remote sensing data, surveys of households, and
ecological data. The creation and use of these sophisticated
data sets raises many issues. For example, more and more of
our data are geocoded. This raises serious questions regarding
data confidentiality. How do researchers maintain the usability
of data while protecting confidentiality when the identifying
variables also are variables in the analysis? Research in this
area lends itself to potential advances in the social, behavioral,
and economic sciences, computer science, and the
mathematical sciences.
CDI Theme: From Data to Knowledge.
Example
The introduction of cyberinfrastructure into formal and informal
learning environments is already beginning to provide learners
at all levels (K to grey) with the skills and literacies needed to
operate effectively in those environments. In order to take full
advantage of the opportunity to learn in these environments,
their design must be based on our best understanding of human
cognitive and interactive styles and capacities. That
understanding, in turn, can be sharpened considerably by the
data now becoming available from observations of students and
teachers interacting with each other and with the cyberenvironment.
CDI themes: Data to Knowledge, Virtual Organizations.
Domains: human-computer interaction, cognitive science,
developmental and learning sciences.
Example
High school teachers and students can explore science through
a virtual laboratory that gives them access to sophisticated
modeling and simulation systems. Impacts of global
phenomena such as climate warming, and local ones such as
earthquakes in susceptible communities, can be investigated.
They can also participate in simultaneous virtual experiments
with classes at remote locations, underscoring how actions in
one region impact another. This innovative approach to science
education depends on breakthroughs in secure virtual
organizations for collaboration and shared control, models and
simulations of natural and built complex systems that are
accessible in real time and can be used and understood by
students, and interdisciplinary approaches to complexity that
help the public understand the relevance of science to daily life.
Themes: Virtual Organizations, Complexity.
Domains: education.