Delphi Frequently Asked Questions (FAQs)
To help provide timely information about all aspects of the program, this page is updated periodically in response to questions from potential performers.
Full information about Delphi and the application process is in the solicitation on SAM.gov. Ask questions via the ARPA-H Solutions site linked below. Please note, you will first need to sign-in or register an account to submit a question.
Teaming & Forming Teams
Delphi anticipates that teaming will be necessary to achieve the goals of the program. Prospective performers are encouraged to form teams with varied technical expertise to submit a proposal to the Delphi ISO. To facilitate this process, we have created a teaming page where prospective performers can share their profiles and learn more about other interested parties.
Yes
UARCs are eligible to participate as prime or sub-performers.
While we encourage team members to be identified prior to solution summary submission, it is not a requirement.
The expectation is that they would be involved starting day 1 to ensure all processes are scalable.
Teams should be formed as soon as possible. Teams must be fully formed by the proposal due date.
All 4 TAs are required.
They can be the same or different people.
Solution Summary Submissions
All templates will be provided with the final ISO posted to SAM.gov.
As per the draft ISO announcement on SAM.gov, the program won't be accepting solutions summaries until after the full solicitation posts. Once the full ISO posts to SAM.gov, you will see an option on the Solutions Portal to submit a Solution Summary.
As quickly as possible.
General Questions
Intellectual property requirements and expectations will be outlined in the final ISO.
There is not a required minimum calendar effort for the lead PI, but time proposed must be sufficient to lead the proposed work.
ARPA-H prioritizes work that will be conducted in the US, but non-US entities are eligible to participate as either prime or sub-proposers
Technical Questions
Any disease that can be tracked by the defined release of hormones or other biological markers, can be proposed as long as sufficient justification is provided as noted in the ISO. Proposers must identify the primary use case of interest, including the disease and the biomarker(s) they intend to detect, the sensing modality, and the frequency of sensing for development and integration with the TA1 chiplet. Sufficient justification should include published studies and does not include basing the entire proposal solely on unpublished data.
The final device must be in a wearable form worn on the skin or a minimally invasive ingestible.
Delphi envisions chiplet-based sensors that are easily manufactured in clean-room environments, are shelf-stable, low cost, and scalable. Within these constraints, we are open to creative proposed solutions.
Yes, this dual-track approach supports both advanced medical devices (minimally invasive wearables and ingestibles) and non-invasive wellness-focused wearables that contact, but not breach the skin. You can choose either as long as sufficient justification is provided and ALL TA’s are addressed as noted in the ISO.
No, the non-invasive track centers on wearables that do not breach the skin’s surface, targeting the health and wellness space. Devices in this category are not regulated by the FDA, and human factors testing at the conclusion of development will evaluate usability, user experience, and behavioral impact. The minimally invasive track encompasses devices designed to penetrate the skin or be ingested, aligning with traditional medical device pathways. These devices are subject to FDA oversight and will undergo rigorous human clinical trials to confirm safety and efficacy.
At program kickoff, discuss with your team’s TA owners (especially TA1) and TA1 owners from other teams. This will be an ongoing interaction, testing chiplet ecosystem interoperability throughout all phases. The goal is to enable seamless interoperability off-the-shelf chiplets from diverse vendors, ensuring absolute compatibility across components and importantly, across the different teams.
No, proposals should present a scalable and cost-effective strategy for packaging chiplets. The expectation is that companies with solid and established expertise in chiplet design, manufacturing, and packaging, or academic institutions closely partnered with such industry leaders will develop optimal packaging.
In vitro and in vivo models should be chosen based on current literature and TA. The proposers must justify their selection and approaches.
As long as they do not interfere or conflict with metrics already specified in the ISO, proposers are encouraged to define additional milestones and metrics to help illustrate progress towards the program goals.
The “Wet” interface encompasses liquid, gaseous, or other interfaces to ensure biomarker sensing. The use of “Wet” was not to restrict to liquid sampling, only to describe the sensing portion open to the body environment.
No. Devices must detect at least one biomarker that is a hormone, cytokine or other low concentration molecular marker. Sensors can be multi-modal with as many sensors as team desires or would be most compelling.
Teams should submit potential analytes that show the breadth of the sensor and its capabilities. Detection does not require different sensing modalities if your technology can detect a breadth of analytes as sufficiently low concentrations.
Please refer to the ISO for specifics. This is discussed under section TA2: “WET” BIOSENSING CHIPLETS THAT ARE MODULAR, INTERCHANGEABLE ON THE MAIN CHIPLET SUBSTRATE AND SPECIFIC FOR THE SELECTED USE CASE.
Either indication is acceptable. Teams should justify choice of application.
We are interested in real biological information. The sensors must transit with the patient either as wearable or ingestible.
No.
If this is needed as a gold standard for the clinical trial, that is acceptable, but the primary device must be wearable or ingestible.
No, but clearly sensors with lower levels of detection will be preferred.
The inclusion and adaptation of current computational capabilities as chiplets is within scope. However, the focus is not on furthering the development of computational components other than those necessary to meet the broader program goals. This is also an opportunity for cost share.
It is within scope.
We default to the FDA’s classification of minimally invasive.
All minimally invasive approaches are acceptable. However, techniques that minimize patient discomfort will be preferred.
The target is 7 days for continuous measurement. The performers can determine their own charging that minimizes burden on the patient.
Depends on the chosen indication and track, teams should justify choices.
Teams should demonstrate prior experience with scaling and sensing and be able to provide a clear path to scalable production that meets program timelines.
It applies to the overall signal.
We are agnostic to the simulated media. The best approach should be proposed and justified.
We are open to formats that minimize burden on the patient. Please choose the most discrete and effective for your device.
Yes, teams may use existing chiplets with justification on how these chiplets will allow them to meet program goals.
Yes.
The goal of the effort is to establish a chiplet ecosystem. We appreciate that some teams may have already integrated into an ASIC, however the goal is to establish an ecosystem for further development, therefore chiplets are required.
They must be integrated into a swallowable capsule.
Yes, the complementary sensors need to be addressed within the chiplet architecture.
Proposers must describe how signals will be exported from the body to a smartphone or other receiver. These may be included under TA1.
No.
Yes
The goal is to maximize compatibility with established chiplet solutions.
Only to the hub device.
ARPA-H is open to the teams proposing a remix, but interoperability must be demonstrated within the first phase of the program.
The metric will remain the LoD in a simulated clinical sample.
Proposers should focus on a single initial indication. Additional indications can be addressed through table of alternate targets for the Phase 2 remix.
Phase 2 is expected to focus on preclinical (animals). Phase 3 will focus on clinical (humans).
Software is auxiliary and may be budgeted for.
We expect development during the program. The end state should be industry-ready solutions.
All these aspects are important for a viable sensor and teams should plan accordingly.
Over the first 2 months of the program, we will convene multiple working groups to come to a consensus on the common interface. The proposal should justify the team’s choice of interface and packaging substrate.
Phase I clinical trials are for safety. However, we expect the information gleaned will result in clinical utility.
As long as the minimally invasive device that penetrates the stratum corneum, is not permanently implanted in the body and maintains the form, fit, and function of a wearable, the depth of skin penetration is not a limiting factor.
The “wet” interface contact can be with a liquid but can also include gaseous phase from skin, sweat, breath, eyes, etc. Indeed, transduction without physical contact with a biological fluid is allowed if it can sense the target soluble factors at an appropriate limit of detection.
To ensure interoperability, during Phases 2 and 3, teams will need to generate sensors for new analytes that may have different requirements for power, memory, or computation, as determined by the program manager. In this scenario, use of other teams’ proven solutions, rather than redevelopment, would be highly efficient and this proof-of-concept will serve as a benchmark for modularity, reusability, and performance.
The wearable device should be kept at a minimal size while still being compatible with mainstream ergonomics. Furthermore, proposals must focus on efficient designs that balance the required extent of the biological marker tracking with the demand on the worn device (e.g., power, space, timing).
Budgets and Contracting
All team members regardless of their organization must comply with the HHS salary cap.
Direct Labor (fully burdened) should be the total dollar amount for your proposed labor. Labor hours should be the total number of estimated labor hours to be performed by all individuals working on the project.
Yes, current HHS salary cap limits apply to all proposals submitted to the Delphi solicitation.
Subawardees do not need to complete separate BOEs and should be included on the prime’s BOE. Further details, including a template, will be provided in the final ISO.
ARPA-H does not reimburse FDA for the clinical trial. Yes, teams should budget for all engagements.
Yes
No. All team members are expected to propose a budget that is appropriate for the tasks included in the technical proposal.