About the AWS Impact Computing Project at the Harvard Data Science Initiative
The AWS Impact Computing Project at the Harvard Data Science Initiative (HDSI) is an alliance with Amazon Web Services (AWS) aimed at reimagining data science and high performance computing to identify potential solutions for society’s most complex challenges.
The three-year effort will support faculty-led research projects across Harvard. Awards to faculty will be made through sponsored research agreements facilitated by the HDSI and Harvard’s Office for Technology Development.
Summary of Opportunity
The HDSI seeks Letters of Intent (LOIs) from faculty interested in pursuing research that will lead to actionable insights into mitigating the adverse environmental, societal, financial, and health impact of climate change as well as increasing resilience. Funding of up to $380,000 total direct costs (maximum three-year term) will be made available for a small number of faculty-led projects on this topic, and HDSI encourages research proposals having smaller budgets and shorter duration and will look to fund a mixture of project sizes. The deadline to submit a 500-word LOI is August 25, 2023.
Projects should identify complex, large-scale, or novel computing or analysis needs, and priority will be given to projects having the potential to drive real-world solutions.
Who is Eligible to Apply for Funding?
This program is open to individuals who (i) hold a faculty appointment at a Harvard school; and (ii) have principal investigator rights at that school. (Please note: Harvard Medical School faculty must hold a faculty appointment with PI rights in one of HMS’s Quad-based, preclinical departments).
Faculty may submit multiple LOIs.
Introduction and Purpose
The adverse impact of climate change is escalating in rapid fashion, and every facet of society and the environment is being impacted. The cascading effect, triggered by the warming of our planet, has become a global crisis with far-reaching consequences for our health and survival.
The purpose of this funding opportunity is to support faculty-led research projects that contribute to our understanding of climate change for example by developing data, data science methodologies, and their tailored applications for: 1) identifying key drivers of climate change, 2) developing effective climate change mitigation strategies; 3) addressing inequities, vulnerabilities; 4) informing adaptation efforts; and 5) quantifying current and future impacts,
The following are examples of research questions that are of interest. These are illustrative only and we welcome LOIs that propose areas of inquiry that fall outside those described below:
Understanding the causes, downstream effects, and interactions underlying climate change.
- Climate evolution and modeling: develop and apply advanced data science methodologies to understand the historical evolution of climate.
- Environmental observation, remote sensing, and resource monitoring: use high performance computing and data science methodologies to assess and monitor, for example, rainforest changes, oceanic changes, and climate and atmospheric impacts to biodiversity.
- Data platforms: develop data platforms and statistical methods to integrate large, heterogeneous datasets for climate, pollution, environmental observation, and sensor fusion.
Data science methods and models
- Data Science Methods to characterize the effectiveness of climate change mitigation and adaptation strategies (e.g., de-carbonization and carbon neutral approaches), predict the course of future climate change, and/or contextualize and anticipate economic and social impact of deployment of smart technologies.
Interventional modeling
- Building design: Leverage computational simulation, building performance evaluation, and design decision support to advance energy efficient building design, materials, and regulations.
- New materials: advance and deploy computational and machine learning methods to accelerate the discovery of new materials that will revolutionize energy storage and conversion, heating and cooling systems, and batteries.
- Biomimetic properties: apply data science methodologies to identify and optimize biomimetic properties in synthetic materials.
Conducting studies on climate change requires analytical sophistication because many environmental and societal factors interact. For example, to mitigate the consequences of the climate crisis we need to understand which factors are key drivers of vulnerability and this can involve high dimensional data.
Developing relevant data science methodologies
- Machine learning: Development and implementation of Machine Learning (ML) methods to identify vulnerable populations and geographies with respect to co-occurring factors.
- Causal inference: Development of causal inference methods that allow the study of the simultaneous causal effects of several concomitant factors or disentangle the causes for confounders in massive spatio-temporal data.
- Synthetic data and validation: Development of synthetic data sets that are representative of vulnerable populations and allow validation of new methods for causal inference and ML.
Submission Process and Requirements
Applicants should submit a 500-word LOI by August 25, 2023.
LOIs will be reviewed by a joint AWS-HDSI review panel and successful submissions will be invited to submit full proposals by October 15th. Funding is expected to begin in early 2024. Prior to full proposal submission, HDSI and AWS will confer with faculty leads and potentially suggest revisions to timelines and scope of proposed projects.
How to Submit an LOI
Please submit your LOI as a PDF to datascience@harvard.edu no later than 11:59 p.m. on August 25, 2023 with the subject line “AWS CLIMATE LOI”. Receipt of your letter will be acknowledged within twenty-four hours. If you do not receive an acknowledgement, please email kevin_doyle@harvard.edu.
LOIs must include the following:
- Title of proposed project
- Names and affiliations of submitting team (PI and any co-PIs)
- Statement of real-world problem that the project addresses
- Description of proposed research questions.
- Description of implications for real-world solutions that may arise from the research, e.g., scalable policy solutions. We especially encourage proposals that identify stakeholders/communities that will benefit from the research outcomes.
- Description of any requirements for large-scale data science, compute-driven modeling, simulation, forecasting, and analytics. Projects should identify complex, large-scale, or novel computing or analysis needs.
- Budget and timeline – this should include initial, direct cost estimates only.
Examples of eligible expenses include:
- Personnel, such as postdocs, research staff, graduate students, undergraduate students
- Travel (domestic and international)
- Acquisition of datasets
Graduate student tuition is NOT eligible for funding.
AWS will make available AWS Cloud Credits to support funded research. At this stage, and where possible, please provide an estimate of need.
Please direct any questions to the HDSI Scientific Director, Lawrence Weissbach at lawrence_weissbach@harvard.edu.