Partnerships within the CISE Directorate and across NSF · • Computer-Assisted Programming for Heterogeneous Architectures (CAPA) • Information-Centric Networking in Wireless

Meghan Houghton, PhDSenior Advisor for Strategic EngagementsComputer and Information Science and EngineeringOffice of the Assistant Director

Partnerships within the CISE Directorate and across NSF

National Science Foundation

“To promote the progress of science; to advance the national health, prosperity, and welfare; to secure the national defense...”

Robotics & Manufacturing

Image Credit: Eliza Grinnell/Harvard SEAS

NSF/CISE programs address national priorities

Advanced Cyberinfrastructure

Image Credit: Texas Advanced Computing Center

Computer Science Education

Image Credit: Calvin Lin, University of Texas, Austin

Smart Communities

Image Credit: US Ignite Image Credit: WINLAB, Rutgers University

Advanced Wireless Research

Image Credit: Photo by Christa Neu; Lehigh University

Quantum Information Sciences

Image Credit: CCC and SIGACT CATCS

Big Data & AI

Image Credit: ThinkStock

Cybersecurity

The value-add Within CISE Across NSF

Outline

Three Primary Objectives:• Deepen and grow research and

innovation;• Make available research

infrastructure; and • Develop the workforce of the

future.

Prescription 3: Establishing a More Robust National Government-University-Industry Research Partnership

Industry & Private Sector

Federal Agencies

International Agencies

Local and State

Government

Universities

Foundations & Non-profits

NSF partners with many stakeholders

The value for NSF and the American research ecosystem:

leverage resources to grow innovations that address real-

world problems; accelerate the translation of discoveries to

deployment; and build research and workforce

capacity.

NSF offers: The industry ROI is…NSF’s gold standard process: soliciting, reviewing, and funding academic research

Access to state-of-the-art, pre-competitive research and a nationwide network of researchers

NSF’s STEM education and workforce development activities

Hiring opportunities for additional well-trained, diverse computer scientists and engineers

NSF offers a range of partnership mechanisms to leverage expertise and resources

Strategically and synergistically advance areas of interest to industry and NSF

Partnerships: the value proposition

Lower Risk Higher

Industry

Higher Market-relevance Lower

Shorter-term Length of investment Longer-term

Lower Foundational/Basic Higher

NSF

NSF’s unique role in supporting R&D

7

The value-add Within CISE Across NSF

Outline

Typical model for each joint solicitation: Total investments: $6-8 million totalFunding ratio: 1:1 NSF:IntelAwards: ~2-6 awards, $500,000-$3 million used over 3 years

Joint Solicitations:• Cyber-Physical Systems Security and Privacy (CPS-Security)• Visual and Experiential Computing (VEC)• Computer-Assisted Programming for Heterogeneous Architectures

(CAPA)• Information-Centric Networking in Wireless Edge Networks (ICN-WEN)• Foundational Microarchitecture Research (FoMR)

Image Credit: CCC and SIGACT CATCS

Image Credit: NSF

Joint funding through an NSF solicitation: Supporting research and researchers Intel example

Model:Partner investment: $3 million in-kind resources committed by each partner

NSF BIGDATA program: Approximately $24 million annually

NSF Awards: ~25-33 awards annually, $600,000-$2 million, 3-4 years

Cloud Resources: Minimum of $100,000 up to a 70:30 split NSF budget: Cloud allocation

Image Credit: NSF

In-kind support through an NSF solicitation: Increasing access to infrastructure and workforce development• Partnership supports large-scale experimentation using cloud resources

and increases access to the infrastructure across institutions.

• Promotes capacity building in the latest technologies for the future workforce.

• FY 2017 BIGDATA program: three major cloud service providers.

• Dear Colleague Letter encouraging participations of cloud service

providers in CISE research.

• FY 2018 BIGDATA program: one additional cloud provider.

$100M public-private investment to create four city-scale testing platforms to enable and accelerate fundamental wireless research going beyond 5G§ $50M CISE investment over 7 years§ $50M Industry Consortium investment from >25

networking vendors, device manufacturers, and wireless carriers

PAWR Project Office managed by:

An NSF-led public-private partnership:Platforms for Advanced Wireless Research (PAWR)

11

Partnering with an international agency: Japan Science and Technology Agency (JST)

• Researcher exchanges at NSF- and JST-Sponsored PI Meetings• Joint Solicitation and Dear Colleague Letter (DCL)

• NSF-JST: US-Japan Big Data and Disaster Research (BDD) Program

• NSF and JST support for travel to build partnerships based on existing grants DCL

• Joint Symposia• JST-NSF International Joint Symposium on Big Data, AI, IoT and Cybersecurity for a New Society, 2016, Tokyo, Japan

• US-Japan Workshop Enabling Global Collaborations in Big Data Research, 2017, Atlanta, GA, US• JST-NSF International Joint Symposium: Challenges for the Future –

The Frontier of Diverse AI Research, 2017, Tokyo, Japan• JST-NSF-DATAIA International Joint Symposium, 2019, Kyoto, Japan

• Memorandum of Collaboration signed by NSF Director and JST President 2018

Institutional partnerships for cloud use: Cloud Access and Exploring Clouds for Acceleration of Science

§ Investigate viability of commercial clouds for leading-edge computational science

§ AWS, Google, Oracle: initial cloud computing providers

§ Builds on AWS, Google, Microsoft, IBM prior partnership, shifting to a new model

§ Jan 2018 CISE workshop motivates path forward

Transformative science that holds promise for tremendous societal and economic benefit with potential to revolutionize how we discover, work, learn, and communicate

• CISE research programs in Information and Intelligent Systems division ($211M FY18)• Cross-directorate programs:§ NRI-2.0: Ubiquitous Collaborative

Robots§ Smart & Connected Communities§ Smart and Connected Health§ Collaborative Research in

Computational NeuroscienceSensing / Data Acquisition

Massive Data Management

Machine Learning

Modeling

AI Infrastructure

Autonomy Human-AI interaction

CISE and Artificial Intelligence

• AI and Society, with the Partnership on AI • NSF/CISE and SBE, with PAI in FY 2019• Jointly funding early-concept grants for exploratory research

• NSF/DARPA Partnership on Real-Time Machine Learning (RTML)• NSF/CISE and ENG beginning in FY 2019• Parallel, synergistic research programs

• NSF/Amazon Program on Fairness in AI • NSF/CISE with Amazon (anticipated in FY 2020)• Jointly funding research grants

Partnerships related to AI looking forward

PAI Founding Members

15

The value-add Within CISE Across NSF

Outline

Sensing / Data Acquisition

Massive Data Management

Machine Learning Components

Modeling

AI Infrastructure

Autonomy Human-AI interaction

The Future of Work at the Human-Technology Frontier

Two Big Ideas with critical ties to AI at NSF

NSF’s 10 Big Ideas for future investment:Convergence Accelerator

17

Renewing NSF agency reform activityTransform NSF into an agile organization capable of

responding to the evolving landscape so that we can enable the Nation’s continued global leadership in

scientific research and innovation.

18

Note: For FY 2017 and FY 2018, seven of the interagency partnership activities also involved one or more international partners.

PARTNERSHIP ACTIVITIES 2017 2018 2019

Interagency 47 57

Private 7 8

Nonprofit/Foundation 3 4

TOTAL 57 68

Partnerships Agency Priority GoalFY 2018-FY 2019 AGENCY PRIORITY GOAL

“By Sept. 30, 2019, partnerships and/or award actions … will grow by 5%, relative to the FY 2017 baseline…”

Examples Renewing NSF deliverables underway

20

• IT: Proposal Submission Modernization• Streamlining: Our internal clearance process• Partnerships: A partnerships toolkit

Thanks!

[email protected]

Cloud Accessentity

Entity: Works with Cloud providers - “establishing partnerships with the various public cloud computing providers”. Works with research community to design “user training and other support to CISE researchers and educators using cloud computing in their work” and “technical guidance for CISE researchers and educators interested in using public cloud computing platforms.”

CISE Cloud Access Solicitation: the mechanism

Cloud Service Provider 1

Cloud Service Provider N

…

CISE-funded PI 1(needing cloud resources)

CISE-funded PI M(needing cloud resources)

…