June 22 -- The National Science Foundation (NSF), through the Directorate for Technology, Innovation and Partnerships (TIP), is launching a new program on Assessing and Predicting Technology Outcomes (APTO) to assess how investments in science and technology research and development will contribute to specific outcomes for the Nation. The APTO program will support a cohort of projects that will work together to complement each other’s research and development (R&D) efforts on technology outcome models to accurately describe three types of technology outcomes: technology capabilities, technology production, and technology use. These models should be able to predict future as well as past states of technology outcomes. Of particular interest are prediction models that are generalizable across multiple technology areas. The outcome of this work will help assess and evaluate the effectiveness of U.S. R&D investments and generate information that decision makers could use to strategize and optimize investments for advancing long-term U.S. competitiveness into the future.
The APTO program serves the TIP directorate’s need for technology assessment to understand where the U.S. stands—as a whole and in individual regions—vis-à-vis competitiveness in the key technology focus areas named in Sec. 10387 of the CHIPS and Science Act. TIP is interested in answers to the question of which science and technology investments would offer the greatest impact in the key technology focus areas and would be essential to the long-term national security and economic prosperity of the United States. As a key aspect of TIP’s technology assessment activity, the APTO program will bring together multidisciplinary teams to help develop the data, intellectual foundations, and analytics necessary to inform decision making.
The research community has accumulated important insights about the “rate and direction of inventive activity” as an aggregate economic good, and about what decision makers can do to increase the overall production of that good. Meanwhile, industry has immense experience with creating specific technologies and planning how to reach intended technology outcomes over periods of several years. The APTO program aims to expand on this knowledge base spanning academia and industry to better understand and predict the long-term evolution of specific technologies over a period of a few years to decades, and specifically model how intentional, purposeful investments can change that evolution.
APTO will fund research and development of causal models that accurately describe past and future technology outcomes, specifically the capabilities, production, and use of specific technologies. These models should be able to predict likely future outcomes for specific technologies and what intentional investments could reliably change or accelerate those outcomes by correctly capturing the various causal relationships. Building and testing these models will require significant amounts of specialized data gathered from a variety of sources, e.g., historical sources, experimentation, expert elicitation, and others. Data extraction and processing tools may need to be developed as part of that effort.
APTO will support a cohort of projects that will work in collaboration on research and development of Technology Outcome Models and in development/preparation of Data Sets and related Tools. The models and methods produced in APTO will enable assessing and evaluating the effectiveness of R&D investments, and whether they are optimized for advancing long-term U.S. competitiveness, in these technology areas as well as others.
The APTO program is interested in supporting R&D in causal models for assessing and predicting outcomes related to technology capabilities, technology production, and technology use. To that end, the program supports two Lines of Effort (LOE) — one on Technology Outcome Models and the other on Data Sets and Tools. The program will select a cohort of projects which, taken together, are best able to cover the full landscape of technology assessment and prediction across the key technology focus areas listed above.
Individual projects may focus on assessment and prediction of one or more technology outcomes (i.e., technology Capability, Production, Use) and on one or both Lines of Effort (i.e., Technology Outcome Models, Data Sets & Tools). The program will strive to cover the full breadth of areas, as best as possible, across the funded cohort of projects.
The technical goals of the APTO program are methods, with related data sets and accompanying tools, that help to accurately assess and predict three types of technology outcomes:
-- Capabilities refers to the state-of-the-art technical performance of an artifact or method;
-- Production is the quantity of creation of an artifact or adoption of a method with a given level of Capabilities; this could be a global total or the total for different regions, industries, etc.; and
-- Use is the count of instances or volume of utilization of an artifact or method with a given level of Capabilities; this could also be a global total or the total for different regions, consumer groups, etc.
The APTO program will support the development of causal models that accurately describe how different types and scales of investments change technology outcomes, and by how much. For example, consider a model for high-performance computing that could be built with data until 1980 and make predictions on the state of the art of computer chip capabilities in 1990 — how many would be produced and what parts of the world would use them. Importantly, the models should be capable of including descriptions of what cause(s) led to these outcomes and be able to make accurate predictions of the outcomes when the purported cause(s) change. This causal understanding is essential for correctly assessing the effectiveness of previous investments and for identifying how future investments might change future outcomes.
Building and testing technology outcome models will likely require significant quantities of specialized data gathered from a variety of sources including historical data, data from experiments, expert elicitation, and others, to serve as model inputs or to test model outputs. Collecting and preparing these data for use in the models will likely require specialized tools. The APTO program will support two Lines of Effort (LOEs) – Development of Technology Outcome Models and Development of reference Data Sets along with associated Tools:
-- Technology Outcome Models – Causal models that accurately predict technology outcomes and what investments change them, along with accompanying software and a reference to data sets employed; and
-- Data Sets and Tools – Derived and/or newly created data sets used for creation and testing of Technology Outcome Models. May include data on technology outcomes or predictors of those outcomes, along with tools developed for efficiently creating such datasets.
Anticipated Funding Amount: $30,000,000. Estimated Number of Awards: 5 to 20.
Up to a total of $30 million may be awarded annually subject to the availability of funds. Awards may be 1 to 5 years in duration. Awards are expected to have an annual budget of $500,000 to $4 million. Awards of longer than 1 year will be made as cooperative agreements. Awards may be made for 1-year efforts for pilots, planning and preparation for larger subsequent efforts to be submitted to future APTO program solicitations.
For multi-year awards, support of each year beyond the first will be contingent on satisfactory review of progress towards targeted goals. Progress will be presented to NSF in a mid-year check-in and to NSF and/or external reviewers towards the end of each year. This review will determine if subsequent-year funding will be awarded.
https://new.nsf.gov/funding/opportunities/assessing-predicting-technology-outcomes-apto