Frontiers in Research and Infrastructure in CISE

Jeannette M. WingAssistant Director

Computer and Information Science and Engineeringand

President’s Professor of Computer ScienceCarnegie Mellon University

Coalition for Academic Scientific ComputationArlington, VA29 April 2009

NSF

3IBM Research Jeannette M. Wing

4IBM Research Jeannette M. Wing

Stewardship for the Field NSF support as a percentof total federal supportof academic basic research

Foundation-wide Initiative

6IBM Research Jeannette M. Wing

CDI: Cyber-Enabled Discovery and Innovation

• Paradigm shift– Not just computing’s metal tools (transistors and wires) but also our mental

tools (abstractions and methods)

• It’s about partnerships and transformative research.– To innovate in/innovatively use computational thinking; and– To advance more than one science/engineering discipline.

• FY08: $48M invested by all directorates and offices– 1800 Letters of Intent, 1300 Preliminary Proposals, 200 Final Proposals, 36

Awards

Computational Thinking for Science and Engineering

7IBM Research Jeannette M. Wing

Range of Disciplines in CDI Awards

• Aerospace engineering• Atmospheric sciences• Biochemistry• Biophysics• Chemical engineering• Communications science and engineering• Computer science• Geosciences• Linguistics• Materials engineering• Mathematics• Mechanical engineering• Molecular biology• Nanocomputing• Neuroscience• Robotics• Social sciences• Statistical physics

… advances via Computational Thinking

8IBM Research Jeannette M. Wing

Range of Societal Issues Addressed

• Cancer therapy• Climate change• Environment• Visually impaired• Water

CISE

10IBM Research Jeannette M. Wing

Core and Cross-Cutting Programs

CNS IISCCF

Core Core

•Algorithmic F’ns•Communications & Information F’ns•Software & Hardware F’ns

• Human-Centered • Information Integra- tion & Informatics• Robust Intelligence

• Computer Systems• Network Systems

• Infrastructure• Education & Workforce

Core

Cross-Cutting• Cyber-Physical Systems• Data-intensive Computing• Network Science and Engineering• Trustworthy Computing

Plus many many other programs with other NSF directorates and other agencies

11IBM Research Jeannette M. Wing

Others

• Joint with other directorates and offices– CISE + BIO + SBE + MPS: Computational Neuroscience (with NIH)– CISE + EHR: Advanced Learning Technologies– CISE + ENG: Cyber-Physical Systems, Multi-core (with SRC)– CISE + MPS: FODAVA (with DHS), MCS– CISE + OCI: DataNet– OCI + CISE + ENG + GEO + MPS: PetaApps– Creative IT (co-funding with other directorates)

• Activities with other agencies, e.g., DARPA, DHS, IARPA, NGA, NIH, NSA• Partnerships with companies

– Google+IBM, HP+Intel+Yahoo!: Data-Intensive Computing– SRC: Multi-core

• Research infrastructure: CRI, MRI• …

Please see website www.cise.nsf.gov for full list.

12IBM Research Jeannette M. Wing

Trustworthy Computing

• Trustworthy = reliability, security, privacy, usability

• Deepen and broaden Cyber Trust

• Three emphases for FY09– Foundations of trustworthy

• Models, logics, algorithms, metrics– Privacy– Usability

Data Intensive Computing

14IBM Research Jeannette M. Wing

How Much Data?• NOAA has ~1 PB climate data (2007)• Wayback machine has ~2 PB (2006)• CERN’s LHC will generate 15 PB a year (2008)• HP is building WalMart a 4PB data warehouse (2007)• Google processes 20 PB a day (2008)• “all words ever spoken by human beings” ~ 5 EB• Int’l Data Corp predicts 1.8 ZB of digital data by 2011

640K ought to be enough for anybody.

Slide source: Jimmy Lin, UMD

15IBM Research Jeannette M. Wing

Convergence in Trends

• Drowning in data

• Data-driven approach in computer science research– graphics, animation, language translation, search, …, computational

biology

• Cheap storage– Seagate Barracuda 1TB hard drive for $90

• Growth in huge data centers

• Data is in the “cloud” not on your machine

• Easier access and programmability by anyone– e.g., Amazon EC2, Google+IBM cluster, Yahoo! Hadoop

16IBM Research Jeannette M. Wing

Data-Intensive ComputingSample Research Questions

Science– What are the fundamental capabilities and limitations of this paradigm? – What new programming abstractions (including models, languages,

algorithms) can accentuate these fundamental capabilities?– What are meaningful metrics of performance and QoS?

Technology– How can we automatically manage the hardware and software of these

systems at scale?– How can we provide security and privacy for simultaneous mutually

untrusted users, for both processing and data? – How can we reduce these systems’ power consumption?

Society– What (new) applications can best exploit this computing paradigm?

17IBM Research Jeannette M. Wing

Data-Intensive Computing Infrastructure for CISE Community

• Google + IBM partnership announced in February 2008– Access to 1600+ nodes, software and services (Hadoop, Tivoli, etc.)– Available to entire community– Cluster Exploratory (CluE) seed program– April 23, 2009: Press release on CluE awards to 14 universities

• http://www.nsf.gov/news/news_summ.jsp?cntn_id=114686&org=NSF&from=news

• HP + Intel + Yahoo! + UIUC cluster announced in July 2009– 1000+ nodes– Bare machine, not just software (Hadoop) accessible– Hosted at UIUC, available to entire community

• Other companies welcome!

Cyber-Physical Systems

19IBM Research Jeannette M. Wing

Smart Cars

Lampson’s Grand Challenge:

Reduce highway traffic deaths to zero.

[Butler Lampson, Getting Computers to Understand, Microsoft, J. ACM 50, 1 (Jan. 2003), pp 70-72.]

Cars drive themselvesCredit: PaulStamatiou.com

A BMW is “now actually a network of computers”

[R. Achatz, Seimens, Economist Oct 11, 2007]

Dash Express:Cars are nodes in a

network

Credit: Dash Navigation, Inc.

Smart parking

20IBM Research Jeannette M. Wing20IBM Research Jeannette M. Wing

Embedded Medical Devices

infusion pump

pacemaker

scanner

Credit: Baxter International

Credit: Siemens AG

21IBM Research Jeannette M. Wing21IBM Research Jeannette M. Wing

Sensors Everywhere

Sonoma Redwood Forest smart buildings

smart bridges

Credit: MO Dept. of Transportation

Credit: Arthur Sanderson at RPI

Hudson River Valley

22IBM Research Jeannette M. Wing

Robots Everywhere

At work: Two ASIMOs working together in coordination to deliver refreshments

Credit: Honda

At home: Paro, therapeutic robotic sealCredit: Paro Robots U.S., Inc.

At home/clinics: Nursebot, robotic assistance for the elderly

Credit: Carnegie Mellon University

At home: iRobot Roomba vacuums your house

23IBM Research Jeannette M. Wing23IBM Research Jeannette M. Wing

Assistive Technologies for Everyone

brain-computer interfaces of today

memex of tomorrow

Credit: Dobelle Institute

Credit: Emotiv

Credit: Paramount Pictures

24IBM Research Jeannette M. Wing

Cyber-Physical SystemsSample Research Challenges

Science• Co-existence of Booleans and Reals

– Discrete systems in a continuous world• Reasoning about uncertainty

– Human, Mother Nature, the Adversary

Technology• Intelligent and safe digital systems that interact with the

physical world• Self-monitoring, real-time learning and adapting

Society• Systems need to be unintrusive, friendly, dependable,

predictable, …

25IBM Research Jeannette M. Wing

A Model for Expediting Progress

FundamentalResearch

auto

finance

civil

aero

medical

chemical materials

energy

IndustryGov’t (e.g., military)

IndustryGov’tAcademia

AcademiaGov’t (NSF, NSA, NIH, DoD, DOE, …)

transportation

Sectors

Network Science and Engineering

27IBM Research Jeannette M. Wing

1999

Our Evolving Networks are Complex

19801970

28IBM Research Jeannette M. Wing

1999

Our Evolving Networks are Complex

19801970

29IBM Research Jeannette M. Wing

1999

Our Evolving Networks are Complex

19801970

30IBM Research Jeannette M. Wing

Network Science and Engineering

• Fundamental Question: Is there a science for understanding the complexity of our networks such that we can engineer them to have predictable behavior?

• Deepen and broaden research agenda of original GENI concept

• Includes CISE’s current networking programs: SING, FIND, NGNI

31IBM Research Jeannette M. Wing

Network Science and EngineeringSample Research Challenges

- Understand emergent behaviors, local–global interactions, system failures and/or degradations- Develop models that accurately predict and control network behaviors

- Develop architectures for self-evolving, robust, manageable future networks- Develop design principles for seamless mobility support- Leverage optical and wireless substrates for reliability and performance- Understand the fundamental potential and limitations of technology

- Design secure, survivable, persistent systems, especially when under attack- Understand technical, economic and legal design trade-offs, enable privacy protection- Explore AI-inspired and game-theoretic paradigms for resource and performance optimization

Science

Technology

SocietyEnable new applications and new economies, while ensuring security and privacy Security, privacy,

economics, AI, social science researchers

Network science and engineering researchers

Understand the complexity of large-scale networks

Distributed systems and substrate researchers

Develop new architectures, exploiting new substrates

32IBM Research Jeannette M. Wing

Network Science and EngineeringSample Research Challenges

- Understand emergent behaviors, local–global interactions, system failures and/or degradations- Develop models that accurately predict and control network behaviors

- Develop architectures for self-evolving, robust, manageable future networks- Develop design principles for seamless mobility support- Leverage optical and wireless substrates for reliability and performance- Understand the fundamental potential and limitations of technology

- Design secure, survivable, persistent systems, especially when under attack- Understand technical, economic and legal design trade-offs, enable privacy protection- Explore AI-inspired and game-theoretic paradigms for resource and performance optimization

Science

Technology

SocietyEnable new applications and new economies, while ensuring security and privacy Security, privacy,

economics, AI, social science researchers

Network science and engineering researchers

Understand the complexity of large-scale networks

Distributed systems and substrate researchers

Develop new architectures, exploiting new substrates

33IBM Research Jeannette M. Wing

NetSE Research Agenda:Background

• In January 2008, the CCC charged the NetSE Council Working Committee led by Ellen Zegura to develop a community-based process that would lead to a comprehensive agenda for the science and engineering of networks.

• Three workshops and two meetings, held between June and October 2008, elaborated the NetSE research space.

• A full draft of the agenda is now circulating among workshop co-chairs, CCC and NetSE members.

• The draft will be posted soon on the CCC website for comment.

July 29, 30, 2008: Science of Network Design, LA, CAOrganizers: John Doyle, Cal Tech; John Wroclawski, USC

July 29, 30, 2008: Science of Network Design, LA, CAOrganizers: John Doyle, Cal Tech; John Wroclawski, USC

July 31, Aug. 1, 2008: Behavior, Computation and Networks in Human Subject Experimentation, Del Mar, CAOrganizers: Michael Kearns, Penn; Colin Camerer, Cal Tech

July 31, Aug. 1, 2008: Behavior, Computation and Networks in Human Subject Experimentation, Del Mar, CAOrganizers: Michael Kearns, Penn; Colin Camerer, Cal Tech

August 17,18, 2008: Network Design and Engineering, Seattle, WAOrganizer: Ellen Zegura, GA Tech

August 17,18, 2008: Network Design and Engineering, Seattle, WAOrganizer: Ellen Zegura, GA Tech

September 24, 25, 2008: Network Design and Societal Values, Arlington, VA

Organizers: Helen Nissenbaum, NYU; Dave Clark, MIT

September 24, 25, 2008: Network Design and Societal Values, Arlington, VA

Organizers: Helen Nissenbaum, NYU; Dave Clark, MIT

Ellen Zegura developed an introduction that synthesizes the main themes across the workshops

Ellen Zegura developed an introduction that synthesizes the main themes across the workshops

June 11, 2008: Theory of Networked Computing, Boston, MA Organizer: Joan Feigenbaum, Yale

June 11, 2008: Theory of Networked Computing, Boston, MA Organizer: Joan Feigenbaum, Yale

34IBM Research Jeannette M. Wing

NetSE Research Agenda

GENIScience Plan

Why WeDream of GENI

35IBM Research Jeannette M. Wing

Network Research Infrastructure

• A parallel activity -- community planning for a suite of infrastructure that will support NetSE experiments -- has been underway since 2005.

• Requirements will evolve with the evolving NetSE research agenda.

• In 2007, CISE funded BBN to provide overall management of and engineering for infrastructure planning and prototyping.

36IBM Research Jeannette M. Wing

GENI Spiral 1• We are here

GENI Prototyping Plan

Multiple competing control frameworks,beginnings of an at-scale experiment plane

37IBM Research Jeannette M. Wing

Key goals for Spiral 1 Clusters

• Integrate “vertically”into the cluster control

framework

• Integrate “horizontally”to create end-to-end slices

• Demonstrate (early) integrated prototypesup and stumbling within the next 3-6 months

• . . . and design GENI as we go!

Components

Aggregate AComputer

Cluster

Components

Aggregate BBackbone Net

Components

Aggregate CMetro Wireless

ReferenceDesign

38IBM Research Jeannette M. Wing

Spiral 1 Academic-Industrial Teams

CNRI

39IBM Research Jeannette M. Wing

Current status - GENI Spiral 1Rapid prototyping, integration, and early experiments

Broadband

41IBM Research Jeannette M. Wing

Broadband• Mid-December

– CISE works with a handful of academics, led by Ed Lazowska (CCC) to provide a vision on Cyberinfrastructure for eScience and eLearning for President-Elect Transition Team.

– CISE independently works with staffers and OLPA on Cyberinfrastructure vision

• Mid-February: “Unleashing Waves of Innovation” white paper released publicly to community– http://www.cra.org/ccc/docs/init/Unleashing.pdf– Endorsed by many state and regional network consortia, National Lambda

Rail, Internet 2, etc.– Latest version on CISE AC website

42IBM Research Jeannette M. Wing

4. Colleges and universities serve as neutral territory for open, non-proprietary, unclassified work, fostering partnerships with industry and government.

1. Colleges and universities are innovation incubators.

5. Colleges and universities are community hubs.

Unleashing Waves of Innovation

CSNET NSFNET vBNS

2. College and university applicationsdrive innovation in networking.

3. Colleges and universities have a four-decade proven track record in deploying and managing advanced networking.

Credit: NASA

Credit: Wikipedia

43IBM Research Jeannette M. Wing

Broadband con’td

• Feb 17: NTIA + RUS get $7B in ARRA for start of national broadband strategy; NSF’s ARRA $200M ARI and $300M MRI are potential sources of funding for universities too. – Be on the lookout for calls for proposals for NTIA BTOP and NSF

ARI and MRI programs.• April 24: Secretary Locke (Commerce) and Dr. Bement

(NSF) meet– NTIA+RUS and NSF will continue to work closely on a national

broadband strategy for ARRA and beyond• April-summer: CISE and OCI will work with NTIA+RUS on

joint advisory committee, review process, etc.

Research Ideas in the Works

45IBM Research Jeannette M. Wing

ClickworkersCollaborative Filtering

Collaborative IntelligenceCollective Intelligence

CrowdsourcingHuman-Based Computation

Recommender SystemsReputation Systems

Social CommerceSwarm Intelligence

WikinomicsWisdom of the Crowds

46IBM Research Jeannette M. Wing

Socially Intelligent Computing

• Multiple dimensions– Numbers and types of people– Numbers and types of devices and services– Numbers and types of communications and interactions

• Examples– Individual Memexes, personalized robots, social networks,

Second Life++, human computation

• Question: Can we harness these capabilities to make humans and computers work effectively in harmony, solving problems neither can solve alone?

47IBM Research Jeannette M. Wing

Green IT

IT as part of the problem and IT as part of the solution

• IT as a consumer of energy– 2% (and growing) of world-wide energy use due to IT

• IT as a helper to solve problems– Direct: reduce energy use, recycle, repurpose, …– Indirect: e-commerce, e-collaboration, telework -> reduction travel, …– Systemic: computational models of climate, species, … -> inform science

and inform policy

• Broader context: Sustainability, Energy, Climate Change, Economy, Human Behavior

48IBM Research Jeannette M. Wing

Computer Science and Economics

- Automated mechanism design underlies electronic commerce, e.g., ad placement, on-line auctions, kidney exchange

- Internet marketplace requires revisiting Nash equilibria model

- Use intractability for voting schemes to circumvent impossibility results

- Emphasis on foundational aspects (e.g., algorithms, game theory) but clearly contributions/participation from AI (multi-agents) and systems communities (and hence overlaps a bit with Human-Computer Intelligence and NetSE)

Computer Science influencing EconomicsEconomics influencing Computer Science

Last Word:The Future of Computing is Bright!

50IBM Research Jeannette M. Wing

Drivers of Computing

Science

Society

Technology• What is computable?• P = NP?• (How) can we build complex systems simply?• What is intelligence?• What is information?

J. Wing, “Five Deep Questions in Computing,” CACM January 2008

7A’sAnytime Anywhere AffordableAccess to Anything by Anyone Authorized.

51IBM Research Jeannette M. Wing

Drivers of Computing

Science

Society

Technology• What is computable?• P = NP?• (How) can we build complex systems simply?• What is intelligence?• What is information?

J. Wing, “Five Deep Questions in Computing,” CACM January 2008

7A’sAnytime Anywhere AffordableAccess to Anything by Anyone Authorized.

Thank You!

53IBM Research Jeannette M. Wing

Credits

• Copyrighted material used under Fair Use. If you are the copyright holder and believe your material has been used unfairly, or if you have any suggestions, feedback, or support, please contact: jsoleil@nsf.gov

• Except where otherwise indicated, permission is granted to copy, distribute, and/or modify all images in this document under the terms of the GNU Free Documentation license, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled “GNU Free Documentation license” (http://commons.wikimedia.org/wiki/Commons:GNU_Free_Documentation_License)

• The inclusion of a logo does not express or imply the endorsement by NSF of the entities' products, services or enterprises