NSERC-CSE Research Communities grants

Overview

On this page

• Background
• Objectives
• Context
• Funding value and duration
• Applicants
• Application
• Equity, diversity, and inclusion
• Award and reporting
• Resources

Background

The Natural Sciences and Engineering Research Council of Canada (NSERC) and the Communications Security Establishment (CSE) are partnering to fund Research Communities to conduct unclassified research on cutting-edge technologies in areas of strategic importance to CSE and the Government of Canada. A Research Community is defined as a group of researchers – and their research personnel (i.e., students, post-doctoral fellows, and research professionals) – from multiple Canadian universities working in related domains and sharing NSERC-CSE funds awarded to their project(s). The first Research Communities grants focus on robust, secure, and safe artificial intelligence (AI). CSE is the sole partner organization; no additional partner organizations will be accepted.

CSE is Canada’s foreign signals intelligence agency and technical authority for cyber security and information assurance. CSE is home to the Tutte Institute for Mathematics and Computing (TIMC), where researchers work with government, academia, and industry to tackle scientific challenges related to CSE’s mission. Data science and AI research at TIMC focuses on extracting information from large and complex datasets. Researchers aim to robustly, securely, and safely gather, parse, and extract salient information from those datasets.

Objectives

• Generate new knowledge in robust, secure, and safe AI from either a model-centric or data-centric view.
• Enhance the capacities of Canadian universities to undertake research related to robust, secure, and safe AI.
• Help produce a new generation of data scientists and engineers sensitive to the issues around robust, secure, and safe AI.

Context

In 2019, member countries of the Organisation for Economic Co-operation and Development (OECD), including Canada, adopted the AI Principles for responsible stewardship of trustworthy AI. Article 1.4 of the AI Principles states that “AI systems should be robust, secure, and safe throughout their entire lifecycle so that, in conditions of normal use, foreseeable use or misuse, or other adverse conditions, they function appropriately and do not pose unreasonable safety risk. To this end, AI actors should ensure traceability, including in relation to datasets, processes and decisions made during the AI system lifecycle, to enable analysis of the AI system’s outcomes and responses to inquiry, appropriate to the context and consistent with the state of art. AI actors should … apply a systematic risk management approach to each phase of the AI system lifecycle on a continuous basis to address risks related to AI systems, including privacy, digital security, safety, and bias.”

A typical AI lifecycle includes the following steps; however, this process is not linear, as each step may feed back into the previous steps:

• Scope project
• Collect and process data
• Train and interpret models
• Validate and verify models
• Deploy models
• Operate and monitor model health and data drift

Robust, secure, and safe AI can be approached from the following perspectives: data-centric, model-centric, and people-centric. While the people-centric approach is important, CSE focuses on the mathematical and statistical perspectives; the data-centric and model-centric approaches.

A data-centric approach focuses on the early stages of the AI process, where the emphasis is on improving data quality to achieve the desired robustness. Examples of data-centric approaches include data understanding (including analysis of bias), data validation and cleaning, and data augmentation. There is comparatively little research on methods for improving and/or responding to data quality issues. For this reason, CSE will favour proposals that incorporate data-centric approaches.

A model-centric approach focuses on the latter stages of the AI process, where the emphasis is on developing models that are robust in the presence of low-quality data. Examples of model-centric approaches to robustness and security include detection and handling of out-of-distribution data, anomaly detection, and uncertainty estimation.

Funding value and duration

Individual projects can request either $700,000 (small project) or $1,400,000 (large project) per year over four years from NSERC. Grant funding is provided in equal parts by NSERC and CSE. The aim is to fund one large or two small projects per call. The first Research Communities grant(s) will be awarded in April 2024.

First, applicants will be asked to submit a letter of intent (LOI) that will be screened by NSERC and CSE. Based on the screening process, selected applicants will be invited to submit a full proposal to the Alliance grants program. NSERC will administer Alliance grant applications and subsequent funding. See Application for details.

Funds will be administered according to NSERC’s use of grant funds guidelines, outlined in the Tri-agency Guide on Financial Administration.

Applicants

This first Research Communities grant(s) will support unclassified research focusing on robust, secure, and safe AI. However, the funding is conditional on the applicant or one of the co-applicants, and some of the Research Community members (i.e., professors, students, and post-doctoral fellows) travel to Ottawa to conduct classified research at CSE throughout the grant period. See Classified research for details.

Unclassified research

Canadian university researchers must be eligible to receive NSERC funds. CSE encourages researchers from multiple universities working in related domains to apply as a group to form a Research Community. However, only one application per researcher will be accepted under this call for proposals (as either applicant or co-applicant).

CSE reserves the right to review and approve the list of Research Community members, including collaborators, prior to their engagement with the research team.

Classified research

In collaboration with CSE researchers, the applicant, or one of the co-applicants, and some of the Research Community members will be required to work on classified research at TIMC, located at CSE’s Edward Drake building in Ottawa. However, the Alliance proposal will only describe the activities and the budget related to the unclassified component of the research.

Applicants must ensure that they can respect the following conditions, or CSE reserves the right to reject the LOI or terminate the grant:

• When submitting the LOI, the applicant, or at least one of the co-applicants, must be a Canadian citizen, eligible for a TOP SECRET security clearance, and willing to work in Ottawa at CSE for at minimum two weeks per year. The identified researcher(s) must maintain their security clearance for the entire grant period.
• Within one year of the grant award, the grant recipient must submit the names of additional Research Community members (i.e., professors, students, and post-doctoral fellows) who are Canadian citizens, eligible for a TOP SECRET security clearance, and willing to work in Ottawa at CSE.
  • Students are welcome under CSE’s Student Program and will be expected to complete at least one semester of work at CSE. See Additional details.
  • Post-doctoral fellows and professors are welcome through CSE’s Interchange Program; the length of their visits will be variable. See Additional details.
• Over the grant period, the grant recipient must continue to submit the names of Research Community members (i.e., professors, students, and post-doctoral fellows) who are Canadian citizens, eligible for a TOP SECRET security clearance, and willing to work in Ottawa at CSE as described in the previous bullet.

Application

Step 1 – LOI

To engage CSE in supporting their research project, applicants must submit an LOI via NSERC’s online system using the LOI template by October 18, 2023, 8:00 p.m. (ET).

The LOI must not exceed three pages, excluding references, and must:

• Describe the research team and identify the researcher(s) (applicant and/or co-applicant(s)) who are eligible for a TOP SECRET clearance. No changes in the research team (applicant and co-applicants) composition are permitted after NSERC and CSE review and approve an LOI.
• Outline the proposed research and explain how the project will address one or more specific research objectives and how it will help enable robust, secure, and safe AI. Include a summary of the proposed research project’s main objectives and challenges and the expected outcomes and benefits for Canada.

In addition to a completed LOI template, applicants must provide the following documents and information to NSERC via the online system:

• A preliminary budget (in Canadian dollars) and accompanying justification to cover the direct cost of the proposed research according to eligible expenses listed in the Tri-agency Guide on Financial Administration. The total Amount requested from NSERC can be either $700,000 (small project) or $1,400,000 (large project) per year over four years. These amounts include CSE’s contribution to the grant funding.
• A completed personal data form with CCV attachment (NSERC form 100A) for the applicant and all co-applicants.

Instructions for submitting documents and information to NSERC:

• Log in to NSERC’s online system and select Create a new form 101
• Select Research partnerships programs, then Alliance grants
• For the Proposal type field, select Letter of Intent
• For the Type of call field, select CSE – Research Communities from the drop-down menu as appropriate

NSERC will internally review the eligibility of the applicant and the co-applicants to receive funds from NSERC under the Alliance grant program. The NSERC eligibility criteria for faculty apply. CSE will review the LOI to determine which proposals meet the objectives detailed above.

The LOI will be evaluated against the following criteria:

• Alignment of proposed research project with topic of the call, including consideration of breadth and/or depth of focus
• Likelihood of impact of proposed research project on CSE operations
• Potential for direct collaboration between Research Community members and CSE researchers

Step 2 – Full application (by invitation only)

Based on the review of the LOI, applicants may be invited to proceed with a full application. NSERC will send a letter of invitation to applicants, and only applications that have been invited to submit will be accepted.

The letter of invitation will provide instructions on how to apply to this call under the Alliance grants program. Full applications must be submitted via NSERC’s online system.

Invited university researchers can apply for a four-year Alliance grant. Individual projects can request either $700,000 (small project) or $1,400,000 (large project) per year over four years from NSERC. Grant funding is provided in equal parts by NSERC and CSE. CSE is the sole partner organization; no additional partner organizations will be accepted on the application.

The deadline to submit the full application is January 31, 2024, 8:00 p.m. (ET).

Full applications will first undergo an administrative assessment by NSERC to ensure they are complete and comply with all requirements. Once the administrative assessment is completed, NSERC will conduct a merit assessment of the applications through a selection committee of external peer reviewers with expertise directly related to this call. The merit of the full applications will be assessed against the following four equally weighted evaluation criteria:

Relevance and expected outcomes

• Significance of the intended outcomes and benefits for Canada
• Originality of the research and potential for generating new scientific knowledge
• Extent to which the strategy chosen to apply the research results is likely to achieve the intended outcomes
• Alignment with the objectives and/or research topics of the call

Proposal

• Clarity of the objectives and deliverables; appropriateness of the scope and size of planned research activities to achieve the expected outcomes; quality and feasibility of the research proposal
• Appropriateness of the identified indicators and methods for monitoring progress during the project and for assessing outcomes afterwards
• Appropriateness of, and justification for, the planned expenditures

Project team

• Appropriateness of academic researchers’ expertise, and that found within the proposed Research Community, both for carrying out the planned research activities and for mentoring trainees
• Clarity of the individual contributions to the research effort and steps taken to ensure an effective collaboration in the completion of the research project
• Added value of the proposed Research Community in achieving the desired outcomes

Training plan

• Opportunities for enriched training experiences for research trainees (undergraduates, graduates, post-doctoral fellows) to develop relevant research skills as well as professional skills such as leadership, communication and collaboration
• Consideration of equity, diversity and inclusion in the training plan (assessed by NSERC staff; for guidance, consult the Equity, diversity and inclusion in your training plan web page)

Upon completion of the external peer review, NSERC will determine the merit of the proposal by applying merit indicators based on the assessment of the evaluation criteria by a Selection Committee and establish a list of fundable applications from which CSE will select the one(s) to be funded based on the following additional criteria:

• Alignment of proposed research project with topic of the call, including consideration of breadth and/or depth of focus
• Likelihood of impact of proposed research project on CSE operations
• Potential for direct collaboration between Research Community members and CSE researchers

The information provided in the application is collected under the authority of the Natural Sciences and Engineering Research Council Act. NSERC is subject to the Access to Information Act and the Privacy Act. The information provided is stored in a series of NSERC data banks as described in Information about programs and information holdings.

Equity, diversity, and inclusion

NSERC is acting on the evidence that achieving a more equitable, diverse, and inclusive Canadian research enterprise is essential to creating the excellent, innovative, and impactful research necessary to advance knowledge and understanding, and to respond to local, national, and global challenges. This principle informs the commitments described in the Tri-agency statement on equity, diversity, and inclusion (EDI).

Applicants are encouraged to consider sex, gender and diversity in the project’s research design. Diversity and gender equity should also be considered when developing the research team and considering how to promote and support a variety of forms of mentoring to ensure trainees’ equitable participation and growth. Applicants should refer to the NSERC guide on integrating equity, diversity and inclusion considerations in research for more information.

Award and reporting

NSERC and CSE aim to announce funding decisions by April 2024.

Those awarded grants will be required to provide periodic progress reports on the financial use of grants and on research progress/results, and a final report once the project is completed. Details on reporting requirements, including scheduling, will be provided with the award letter and the terms and conditions of the award.

Resources

• NSERC-CSE Research Community grants terms and conditions of applying for applicants
• NSERC-CSE Research Community grants terms and conditions of award