TRL 3 meaning: proof of concept, and the evidence that counts
TRL 3 meaning explained: experimental proof of concept, how NASA, the EU and India's NTRAF define it, the nine critical NTRAF checks and lab examples.
Part of the guide TRL levels explained: what TRL 1 to 9 really mean, and how to place your technology
TRL 3 means experimental proof of concept: laboratory experiments or analytical studies have shown that the critical function of a technology works as predicted. NASA's exit criterion is documented results validating predictions of key parameters. India's NTRAF lists nine critical checks for TRL 3, including laboratory experiments that verify feasibility and components tested at laboratory scale.
TRL 3 is where a research idea stops being an argument and becomes a result. At TRL 2 the lab file says "we predict". At TRL 3 it says "we tested, and here is the data". The experiment may be small and crude, run on one component in a beaker or on a breadboard, but it shows that the critical function works.
It is also a level that is often overstated in grant proposals. A promising first result is not the same as a validated one. This post explains how NASA, the European Union and India's National Technology Readiness Assessment Framework (NTRAF) define TRL 3, what the nine critical NTRAF checks ask for, what TRL 3 looks like at the bench and how to move on to TRL 4. If you are still working from a concept on paper, start with TRL 2 meaning.
What does TRL 3 mean?
TRL 3 means analytical and experimental critical function or characteristic proof of concept. In plain terms, three things are true:
- The one function the technology depends on has been tested, by experiment, analysis or both.
- The results match the predictions made at TRL 2, within a range you can state.
- The parts have not yet been put together and tested as a working system. That is TRL 4.
The word "critical" matters. A sensor has many functions, but one of them decides whether the idea lives or dies: does the material actually respond to the target? TRL 3 asks you to prove that one first.
How do NASA, the EU and NTRAF define TRL 3?
| Framework | Name or definition | What it says | Evidence asked for |
|---|---|---|---|
| NASA ESTO | Analytical and experimental critical function and/or characteristic proof of concept | Analytical studies place the technology in context; laboratory demonstrations, modelling and simulation validate analytical prediction | Documented analytical or experimental results validating predictions of key parameters |
| EU Horizon Europe | Experimental proof of concept | One line in the General Annexes | Set by each call |
| India, NTRAF | Experimental proof of concept; stage: research to prove feasibility | "Active Research and Development (R&D) is initiated." Work moves beyond paper to experiments, and laboratory studies validate predictions | A proof-of-concept simulation report for the first question, then nine critical checklist items |
Sources: NASA ESTO technology readiness levels, the Horizon Europe General Annexes 2026 to 2027 and the National Technology Readiness Assessment Framework.
NASA also gives a software description for TRL 3: development of limited functionality to validate critical properties and predictions, using non-integrated software components. For a software team, that means the core algorithm runs and gives the predicted answer on test data, but it is not yet wired into the rest of the product.
NTRAF places TRL 3 in the same "research to prove feasibility" stage as TRL 2, and describes TRL 4 to 6, its technology demonstration stage, as the bridge from scientific research to engineering. TRL 3 is therefore the last level that is still mainly research.
TRL 3 at the end of the research stage
What are NTRAF's critical checks for TRL 3?
NTRAF's assessment has two steps at each level. The first asks one question: "Has the proof of concept been demonstrated in a simulated environment?", supported by a proof-of-concept simulation report giving the method, data sources and results. The second is a checklist. Every item marked critical must be answered yes, with documents, or the technology is pegged at TRL 2.
| Critical item in NTRAF | What it looks like in a lab file |
|---|---|
| Predictions of elements or components validated by analytical or experimental studies | A table of predicted against measured values for each key parameter |
| Preliminary system performance characteristics and measures identified and estimated | The numbers that will define success for the full system, with first estimates |
| Laboratory experiments verify feasibility of application | Raw data, method and repeat runs for the critical experiment |
| Predictions of technology capability validated by laboratory experiments | Evidence that the measured performance supports the intended use |
| Cross-technology effects, if any, have begun to be identified | Notes on how one component affects another, such as heat, interference or contamination |
| Performance metrics for the system established | Target values and how each will be measured |
| Scaling studies have been started | A first look at what changes at larger size or volume |
| Individual components tested at laboratory scale | Test records for each part on its own |
| Overall system requirements for the end user's application are known | A short requirements list written from the user's side |
Two items catch researchers out. Scaling studies sound like a later job, but NTRAF wants them started at TRL 3, even if that means one page of calculations. And end-user requirements need someone outside the lab: a clinician, a plant engineer or a farmer. Without that conversation the item cannot honestly be ticked.
What does TRL 3 look like at the bench?
Taking the same three examples as the TRL 2 post one step further:
| Field | TRL 2 concept | TRL 3 proof of concept |
|---|---|---|
| Materials | A pH-sensitive polymer film as a valve in a drug-release patch | Film samples tested at several pH values; swelling and release rate measured and compared with prediction |
| Biotech | A heat-stable enzyme for hot-wash detergent | Enzyme activity measured at wash temperatures in buffer and in a simple detergent base; results within the predicted range |
| Electronics | A nanomaterial humidity sensor for grain storage | A single sensing element on a test board; resistance logged across controlled humidity steps; drift measured over days |
In each case, notice what is missing. The film is not yet inside a patch. The enzyme has not been formulated with the full detergent. The sensor has no housing, power supply or reader. Each component has been shown to work. Nothing has been integrated. That gap is exactly what TRL 4 closes, and moving from TRL 3 to TRL 4 sets out a validation plan for it.
A good TRL 3 result is also repeatable. One lucky run proves very little. Three runs on different days, with the conditions written down, are far more convincing to a reviewer or a licensee.
How much does a TRL 3 study cost?
There is no standard figure. It depends on the field, the equipment already available and how many components need testing.
Illustration with invented figures for a university lab proving the humidity sensor concept above, using equipment it already has:
| Item | Cost |
|---|---|
| Nanomaterial synthesis, three batches | ₹60,000 |
| Test boards and electrodes | ₹40,000 |
| Humidity chamber time, 60 hours at an internal rate | ₹45,000 |
| Characterisation (microscopy and spectroscopy) | ₹75,000 |
| Project assistant, four months | ₹1,20,000 |
| Total | ₹3,40,000 |
The point of the example is the shape of the budget. At TRL 3 most of the money goes on materials, characterisation and people's time, not on hardware. Costs rise sharply at TRL 4 and beyond, when parts must be integrated and tested as a system.
Can you patent at TRL 3?
Often this is a good time to file. Patentability under the Patents Act, 1970 depends on novelty, inventive step and industrial application. Readiness is not a test. Even so, TRL 3 data helps in two ways. It shows the invention works, which supports a specification that enables a skilled person to perform it. And it tells you which embodiments to claim, because you now know what actually performed.
The risk is publication. TRL 3 results are exactly what researchers want to present at a conference. Decide on filing before the abstract goes in, and talk to your IPR cell early. For the steps, see how to file a patent in India.
Building a TRL 3 evidence file
Common mistakes
- Claiming TRL 3 from a literature result. Someone else's experiment on a similar material does not prove your concept. NTRAF wants your own laboratory evidence.
- Testing an easy function instead of the critical one. A sensor that has been shown to be cheap to make, but not shown to respond to the target, is still at TRL 2.
- Skipping the predictions. If you did not write down what you expected before the experiment, you cannot show that the results validate a prediction.
- Reporting one run. Repeatability is part of proof.
- Calling an integrated benchtop rig TRL 3. If several components already work together, you may be at TRL 4, and you should say so.
My view is that TRL 3 is the most useful level to be honest about, because everything that follows is built on it. A modest, well-documented proof of concept is worth more to a licensee than an ambitious claim of TRL 4 that falls apart in due diligence. I am less certain how strictly each funding agency applies the NTRAF checklist today, since agencies adopt it at different speeds, so read the scheme guideline you are applying under.
If you want a structured check of where your work sits, TRL Pro assesses technology readiness from TRL 1 to 9 using Zernyx's own copyrighted framework (ZTRAF), which covers more than 60 technology domains and subdomains.
Common questions
What is the difference between TRL 3 and TRL 4?
At TRL 3 the critical function has been shown to work, often in separate experiments on individual components. At TRL 4 the components are put together and validated as a basic system in the lab. If your parts work on their own but have not yet been tested together, you are at TRL 3.
What evidence proves TRL 3?
NASA asks for documented analytical or experimental results that validate predictions of key parameters. NTRAF asks for nine critical items, including validated predictions, laboratory experiments that verify feasibility, established performance metrics, scaling studies that have started and components tested at laboratory scale.
Is a computer simulation enough for TRL 3?
Sometimes, as part of the case. NASA's TRL 3 wording includes modelling and simulation that validate analytical prediction, and NTRAF's first question asks whether proof of concept has been shown in a simulated environment. For most physical technologies, reviewers still expect at least one laboratory experiment on the critical function.
Which grants fund work at TRL 3?
Proof-of-concept grants. BIRAC's Biotechnology Ignition Grant, for example, gives up to ₹50 lakh over 18 months to take an idea to proof of concept. Check each scheme's current guideline for the readiness level it expects.