A research grant pays you to find out whether something works. A pilot or demonstration award pays you to prove it works outside the lab — at a real site, on real equipment, in front of people who will keep operating it after your engineers go home. Founders who submit a demonstration proposal written like a research proposal lose on the same three points every time: no committed host site, no credible cost share, and no plan for what happens to the hardware when the award period ends.
This money is large, less crowded than the general grant pool, and structured very differently. Here is where it sits and what a proposal must contain to survive review.
Demonstration funding versus research funding
The federal government splits technology money roughly into research, development, demonstration and deployment. Research and early development buys knowledge. Demonstration and deployment buys evidence that the technology performs at scale, at cost, under real operating conditions. That distinction drives almost every difference in how the two are administered.
| Dimension | Research grant | Pilot or demonstration award |
|---|---|---|
| Maturity expected | Concept through lab-validated prototype | Working system, already validated in a relevant environment |
| Cost share | Often none, or a small match | Commonly 20–50 percent of total project cost |
| Deliverable | Data, publications, a report | An operating installation plus measured performance data |
| Third parties | Optional subawards | Host site, utility, agency or hospital must be committed in writing |
| Award size | Tens to hundreds of thousands | Often several million, multi-year, milestone-gated |
If your product has never been built as a full working unit, a demonstration solicitation is the wrong door. Go through the SBIR phases first and come back with a Phase II prototype and test data.
Where federal pilot and demonstration money actually sits
Department of Energy
DOE runs the largest concentration of demonstration and deployment funding in the federal government, spread across its applied energy offices and a dedicated clean energy demonstration office. Typical structures are cost-shared cooperative agreements rather than plain grants, meaning a federal project officer stays involved, milestones are negotiated, and payments follow verified progress. Cost share is usually substantial and can include in-kind contributions from the host site. National laboratory partnerships and voucher programs also sit here.
Transportation
The Federal Highway Administration and the Federal Transit Administration both fund pilot deployments, usually through state departments of transportation or transit agencies rather than directly to a vendor. That matters: your route to the money is often to become the technology partner on an agency application, not the applicant. The FAA funds test-site work for unmanned aircraft on similar terms. If you are building anything that flies, the same logic in BVLOS waiver work applies to a pilot deployment.
Health
NIH funds pilot and feasibility studies through specific mechanisms and through center grants that carry internal pilot pools. AHRQ funds health services demonstrations — how a technology changes care delivery, not whether the device works physically. CMS runs payment and care-model demonstrations that decide whether a product gets reimbursed at all. For medical hardware, a reimbursement pilot can be worth more than the device grant that preceded it. Start with the funding paths in NIH funding for medical device startups before assuming SBIR is the only option.
Defense
DoD buys prototypes through Other Transaction Agreements, which are not procurement contracts and not grants. They are negotiated agreements, often awarded through consortia that a company joins, and they can convert to a production agreement without a new competition if the prototype succeeds. Cost share is common, timelines are short, and the counterparty is a program office with an operational problem, not a review panel. Middle-tier acquisition pathways serve the same purpose for larger efforts.
ARPA-style programs
The advanced research project agencies at DOE, HHS and DoD run on a program manager model. A program manager defines a hard technical target, funds a portfolio against it, and actively manages the teams. These are not always-open solicitations, and the framing of each call reflects one person's thesis. Understanding that changes how you write. So does knowing which agency your work actually belongs to.
Cost share is the line that eliminates most applicants
A demonstration solicitation that requires 20 percent non-federal cost share on a $5 million project means you must produce $1.25 million that is not federal money, and you must document it. The most common ways teams fail here:
- Counting other federal money as match. You cannot. Federal funds do not cost-share federal funds.
- Promising future revenue or investment that has not closed. Reviewers want letters, signed agreements, or board-approved commitments.
- Valuing in-kind contributions with no defensible basis. Donated equipment, staff time and host site labor can count, but only at documented rates with an audit trail.
- Ignoring that unspent cost share is a compliance problem. If you claim it, you owe it, in the same proportion, throughout the project.
Cost share is also where private capital and public money have to be sequenced carefully — see how to combine federal grants with private investment without breaking either one.
The host site is harder to get than the money
Almost every pilot solicitation requires a real place where the technology will operate: a factory, a hospital, a transit yard, a municipal water plant, a farm. Getting that commitment takes months and looks nothing like a sales cycle. The host wants to know who pays for downtime, who insures the equipment, who removes it if it fails, who owns the data it produces, and who takes the call at 2 a.m. Draft a site agreement covering all five before you write the proposal, and attach the letter of commitment. The mechanics are close to closing a first pilot with a business customer, with permitting and public oversight layered on top.
Budget real time for site preparation, environmental review, utility interconnection or review board approval. These sit on the critical path and are outside your control.
What a demonstration budget has to include
Research budgets are mostly labor. Demonstration budgets are labor plus a long tail of things first-time applicants forget: shipping and rigging, site civil work, permits, commissioning, spare parts, instrumentation and data acquisition, operator training, insurance riders, monitoring for the full performance period, and decommissioning or removal at the end. Equipment purchased under a federal award also carries property management rules, including disposition at closeout.
Milestone structure matters as much as totals. Cooperative agreements release funds against verified milestones with a go/no-go decision partway through, so build the schedule so each phase produces something a project officer can confirm. The discipline is described in grant reporting and milestone compliance.
How to position the proposal
Demonstration reviewers are asking a different question than research reviewers: not is this interesting, but will this work at this site, at this cost, on this schedule, and will anyone buy it afterward. Lead with the operating data you already have. Name the host and quantify the site conditions. Show the cost trajectory from the demonstration unit to a commercial unit and say what drives it down. State plainly what happens after the award — who operates the system, who pays for it, what the next installation looks like.
And be honest about whether you want this at all. A demonstration award brings a federal project officer, quarterly reporting, audit exposure and a cost-share obligation. For some companies that is a bargain; for others, the same money from investors costs less in total. The tradeoff is laid out in grants versus equity funding.
Projects House helps hardware teams get a product to the maturity a demonstration program expects — a working system, instrumented and documented. If you are targeting a pilot solicitation and need the engineering behind it, tell us about the project through our contact form.