Technology readiness levels: evidence for industrial adoption

Technology readiness levels describe how far a defined technology has progressed from scientific principles to demonstrated operation. They are evidence categories, not percentages of project completion or guarantees of commercial success. An industrial assessment must identify the component or system being rated, the required environment and the documents supporting the judgement. The nine-level ladder is useful for organising technical evidence; adoption also requires manufacturing capacity, integration, operating economics and a supplier capable of supporting the installation.

Which TRL definition should an assessment use?

Use the definition required by the relevant programme, contract or funding call. NASA's introductory framework describes nine stages, with later levels explicitly referring to flight qualification and mission operation. Industrial applications can use analogous operating environments, but should not silently replace a programme's formal terminology. A furnace, software package and space instrument may need very different evidence even when their assigned level is the same.

The following industrial examples illustrate the progression; they are not formal certifications. NASA's Earth Science Technology Office publishes separate hardware and software descriptions with exit criteria. This matters because a maturity claim should point to test records, configuration and conditions. The number becomes useful only when another reviewer can understand why the available evidence satisfies the chosen definition.

What do TRL 1, 2 and 3 establish?

At TRL 1, observed scientific principles provide a possible foundation. For an industrial sensing concept, this might mean evidence that a material changes an electrical property when exposed to a substance. It need not establish a useful instrument. The relevant question is whether the underlying effect is sufficiently understood to justify investigating an application, rather than whether a factory should already purchase it.

At TRL 2, an application or technology concept has been formulated. An engineering team might describe how that sensing effect could detect contamination in a liquid stream. The concept should identify an intended benefit and plausible operating arrangement, but key assumptions remain untested. A diagram and potential customer interest can strengthen the development case without constituting experimental proof.

At TRL 3, analysis and experiment support a critical function or proof of concept. The sensing team might show that a laboratory sample produces a measurable response at relevant concentrations. The record should distinguish the intended signal from interference and explain what remains unresolved. Demonstrating one physical effect does not establish a complete monitoring system or its ability to survive industrial cleaning.

What changes at TRL 4, 5 and 6?

TRL 4 concerns validation of components or an experimental assembly in the laboratory. In the illustrative sensing project, the sensing element, electronics and data processing could now operate together. Evidence should describe the configuration, calibration and test conditions. If a technician manually performs a step intended to be automatic, that intervention belongs in the record because it defines the demonstration's actual scope.

TRL 5 extends validation into conditions relevant to the intended use. The team might test process-representative temperature, chemical interference and flow variation. Relevant does not simply mean that the apparatus has been moved outside a university. The environment must reproduce the stresses that could prevent the claimed function from working. Missing stresses should remain visible rather than being hidden by the numerical label.

TRL 6 concerns a representative system or subsystem prototype demonstration. For industrial interpretation, the prototype should address the scaling and integration questions that dominate the next step. The sensing installation might run on a representative process loop with normal controls and realistic flow rates. NASA sources use environment terminology somewhat differently at this boundary, reinforcing the need to name the specific framework and its exit criteria.

What evidence supports TRL 7, 8 and 9?

TRL 7 concerns a system prototype in its operational setting. The illustrative sensor might now operate on the intended production line, with genuine disturbances, operating staff and maintenance constraints. A demonstration needs a recorded duration and success conditions. Calling the setting operational does not prove that every shift, product recipe or seasonal condition has been covered.

TRL 8 concerns the completed system being qualified through the required testing. The relevant configuration should be controlled, with the documentation needed for its intended use. For the sensor, this could include an agreed installation configuration, operating procedures and completed acceptance evidence. A significant redesign after qualification requires an assessment of which earlier results remain applicable.

TRL 9 concerns successful operation of the actual system in its intended setting. For an industrial product, useful evidence identifies where, for how long and under which demands the system operated. This supports a mature technical claim for that application. It does not automatically validate a hotter process, a different feedstock or a manufacturing expansion that changes how the equipment is built.

How should a technology readiness assessment be organised?

The US Government Accountability Office's assessment guide treats readiness assessment as an evidence-based process focused on technologies critical to a larger system. For a factory project, define the decision first: funding further research, placing a pilot order and committing to production replacement are different decisions. Identify the critical uncertainties and assemble reviewers with the relevant technical and operating knowledge.

For each assessed element, record the configuration, intended environment, claimed level, evidence reviewed, limitations and next test. Preserve unsuccessful runs and unexplained deviations. A technology portfolio should not average a mature component with an immature one into a reassuring system score; an unresolved component may govern the installation's whole operating risk. Where assessment relies on supplier-owned data, state which results were independently witnessed or reproduced.

Why is a high TRL insufficient for industrial adoption?

NASA's technology assessment and insertion guidance connects maturity with the work needed to incorporate technology into a system. A mature component in a new arrangement can still create integration risk. An industrial buyer must separately examine production quantity, installation interruption, spare parts, software dependencies, service response and responsibility for process failures. The adoption decision therefore requires both technical evidence and a viable operating arrangement.

A bounded pilot should resolve a stated adoption uncertainty. For inspection equipment, agree acceptable missed defects, false rejections and throughput under normal product variation. Reserve some specimens or operating conditions for independent evaluation. The metal additive-manufacturing guide illustrates why process qualification matters, while heat-pump procurement shows how a mature machine can still require extensive site-specific engineering. Readiness is most useful when it directs the next investment in evidence.