Design Intelligence Fostering Formidable Energy Reduction and Enabling Novel Totally Impactful Advanced Technology Enhancements (DIFFERENTIATE) | DE FOA 0002107

DIFFERENTIATE (ARPA-E) Grant Opportunity FAQs

1) What is the DIFFERENTIATE funding opportunity?

DIFFERENTIATE is an ARPA-E (U.S. Department of Energy) research and development program focused on speeding up energy technology innovation by integrating modern machine learning methods directly into the engineering design workflow. The goal is to shorten and improve the iterative design cycle that typically relies on a mix of human intuition, lower-cost modeling, and expensive high-fidelity simulations or experiments.

2) Who is offering this funding?

The program is offered by ARPA-E, which is part of the U.S. Department of Energy (DOE). ARPA-E’s mission is to support high-potential, high-impact energy technologies that are too early-stage or too risky for typical private-sector investment.

3) What are ARPA-E’s broader goals for supporting projects like this?

ARPA-E supports projects to advance national goals including improving U.S. economic and energy security, reducing energy-related emissions (including greenhouse gases), decreasing reliance on foreign energy imports, and strengthening U.S. technological leadership.

4) What type of award instrument is used under this opportunity?

Awards under this announcement are issued as cooperative agreements. That means ARPA-E expects to be actively involved in project execution and management, rather than serving as a hands-off funder.

5) What is the core problem DIFFERENTIATE is trying to solve?

DIFFERENTIATE targets the slow, expensive, and intuition-heavy engineering design loop. In many energy technology areas, teams iterate through conceptual ideas, refine them with cheaper models, and then validate them using expensive high-fidelity simulations or experiments. Because high-fidelity evaluation can be costly and time-consuming, design iterations can bottleneck and progress can be slow. The program aims to compress cycle time by developing ML-enhanced tools that increase engineer productivity and reduce bottlenecks created by optimization and complex tradeoffs.

6) How does the program describe a typical engineering design workflow?

The opportunity describes an iterative loop where teams: (1) start with a clearly stated challenge, (2) propose candidate concepts, (3) refine those concepts using lower-cost models, (4) test/evaluate concepts with expensive, high-fidelity simulations or experiments, and (5) feed those results back into the next iteration.

7) What examples of high-fidelity evaluations are mentioned?

The opportunity mentions examples such as computational fluid dynamics (CFD), finite element analysis (FEA), and system demonstrations as forms of expensive, high-fidelity evaluation that can be part of the design loop.

8) What are the main tool areas DIFFERENTIATE wants to advance?

The program outlines three primary tool areas: (1) machine-learning-enhanced hypothesis generation tools, (2) machine-learning-enhanced high-fidelity hypothesis evaluation tools, and (3) inverse design tools.

9) What are machine-learning-enhanced hypothesis generation tools?

These are tools aimed at improving conceptual design. They are intended to help teams generate and refine promising ideas more quickly than traditional brainstorming paired with basic modeling.

10) What are machine-learning-enhanced high-fidelity hypothesis evaluation tools?

These tools focus on detailed design, where machine learning can reduce the time and cost of high-fidelity solvers and experiments. The description includes approaches like learning surrogate models, improving sampling strategies, and focusing computation on the most informative regions of the design space.

11) What are inverse design tools in the context of this program?

Inverse design tools reverse the typical design approach. Instead of manually proposing designs and then checking performance, inverse design starts from desired performance targets and uses optimization and machine learning to suggest designs that are likely to meet those targets. This can help explore non-obvious solutions that human designers might overlook.

12) Does the program emphasize specific team compositions or collaborations?

Yes. ARPA-E emphasizes that progress in this area requires interdisciplinary teams. The FOA explicitly encourages collaborations that combine mathematicians, operations research analysts, computer scientists, and energy engineers, along with other relevant specialists.

13) Why does ARPA-E emphasize interdisciplinary work for DIFFERENTIATE?

The program notes that many energy technology challenges share common mathematical structures, especially optimization under constraints, uncertainty, and expensive evaluation functions. Because of this, general-purpose ML-enabled design methods can potentially translate across multiple energy R&D domains and accelerate progress broadly.

14) Is the program intended to apply only to a single energy technology niche?

Based on the description, the underlying intent is that improved ML-enabled design methods could be broadly applicable across energy R&D areas, rather than being limited to one narrow technology domain.

15) What is the Funding Opportunity Number (FOA number)?

The Funding Opportunity Number listed for DIFFERENTIATE is DE-FOA-0002107.

16) What CFDA number is associated with this opportunity?

The opportunity is listed under CFDA 81.135.

17) How is the funding categorized?

The listing describes the opportunity as discretionary funding for science and technology and other research and development (R&D).

18) Who is eligible to apply?

Eligibility is described as unrestricted, meaning it is open to any type of entity, subject to any further clarifications in the full FOA.

19) What is the maximum award amount (award ceiling)?

The award ceiling listed is $5,000,000.

20) How many awards does ARPA-E expect to make?

The opportunity notes an expectation of about 7 awards.

21) What was the closing date for concept papers?

The original closing date for concept papers was May 20, 2019.

22) Was there guidance about when to submit relative to the deadline?

Yes. Applicants were encouraged to submit at least 48 hours before the deadline.

23) Where were the full announcement and submission instructions provided?

The full announcement and submission instructions were made available through ARPA-E’s FOA website.

24) What does it mean for applicants that ARPA-E uses a cooperative agreement?

It means ARPA-E expects to be actively involved during the project, including execution and management, rather than simply providing funds and stepping back. Applicants should be prepared for an engaged program management model.

25) What kind of engineering challenge does the FOA use as an example?

The description gives an example of a clearly stated challenge such as achieving very high-efficiency power generation at acceptable cost.