Technology Readiness Assessment

The Program That Was Ready Except for the Part That Mattered

A cislunar propulsion program enters preliminary design review with a technology readiness summary that reassures everyone in the room. The average TRL across the subsystem is 7. The mission schedule shows a path to launch commit two years out with the usual engineering margin. The program’s executives sign off on a procurement package built around that summary, and the industrial partners begin long-lead orders on the assumption that the summary is accurate.

Eighteen months later, the program is slipping. The slip does not originate in the mature subsystems; those are performing as expected. It originates in a single component — a power processing unit running at thermal duty cycles that were never qualified at the full system level, because no ground facility existed that could host the full-power test. The component was listed at TRL 5 in the original assessment, rolled into an arithmetic average with TRL 8 and TRL 9 partners, and the average smoothed the anomaly out of executive visibility. The program was never as ready as the summary said it was. The summary was wrong not because anyone lied but because it averaged instead of measured.

This is the pattern technology readiness assessment was designed to prevent, and the pattern it still fails to prevent when the discipline of its assignment is treated as a formality rather than an analytical act. The TRL scale itself is straightforward. The rigor required to use it honestly is not.

A NASA Tool That Grew Into an International Standard

The TRL concept originated at NASA in the 1970s, where engineers within the agency’s research programs were trying to find a common vocabulary for describing the maturity of candidate technologies under consideration for future missions. Stan Sadin’s 1974 seven-level formulation was the direct ancestor of the scale in use today. The scale was expanded to nine levels in the 1990s by John Mankins and colleagues, whose 1995 white paper gave the framework the definitions that remain largely intact. TRL 1 marked the observation of basic principles; TRL 9 marked a system proven in its operational environment, with every intermediate level describing a specific combination of fidelity of the prototype and fidelity of the test environment.

The scale spread first within NASA, then across the US defense procurement community, then into the European Space Agency and JAXA, and eventually into ISO 16290, which codified TRL definitions for space hardware in 2013 and made the scale a genuinely international reference. The spread was not only semantic. Each adopting agency brought its own procurement and review culture to the scale, which produced a productive tension: ESA’s TRL usage emphasizes environmental qualification more explicitly than early NASA practice; US defense usage added programmatic and integration readiness dimensions. ISO 16290 attempts a harmonization, though practitioners working across agencies still encounter subtle differences in how specific transitions are evidenced.

The scale’s enduring contribution is not its particular level definitions but the discipline it imposes on maturity conversations. Before TRL, technology maturity was discussed in prose, with each program using its own vocabulary and with cross-program comparisons frequently impossible. After TRL, the question “how mature is this?” has a structured answer, and the gap between structured answers and structured requirements can be analyzed.

What the TRL Discipline Actually Demands

The characteristic analytical gesture of technology readiness assessment is the replacement of a summary judgment with an evidence-backed assignment at the subsystem level. The method’s value comes not from the number itself but from three disciplines around the number.

The gap analysis is where the assessment becomes a decision tool. For each element below the target TRL, the analyst estimates the specific testing and validation gaps, the cost and time to advance each level, the risks at each transition, and the facility and expertise requirements. The 5→6 and 7→8 transitions are historically the most expensive and delay-prone, because each involves the step from component-level to system-level environmental demonstration, and ground facilities capable of the latter are often rare and fully booked. A program plan that assumes linear time and cost for TRL advancement is almost always wrong.

A Cislunar Propulsion Module, Read Honestly

Consider the assessment applied to a solar-electric propulsion module for a generic cislunar tug. The decomposition yields four elements: the Hall-effect thruster at the rated power level, the power processing unit, the xenon feed system, and the system-level integration into the tug platform.

The thruster element reads at TRL 6 on verified evidence — laboratory demonstrations at the target power in relevant vacuum conditions, component-level life testing partially completed. The target is TRL 8 by launch commit, a two-level gap. The critical risk is lifetime qualification in the cislunar radiation environment, which requires access to specific radiation test facilities whose scheduling is a known constraint on the program.

The power processing unit reads at TRL 5 with three levels to close by target. Its binding risk is thermal management at the full duty cycle expected in operations. The unit has been demonstrated at duty cycles lower than operational requirements because the ground test facility constraints have prevented full-envelope testing. The 5→6 transition is the most expensive in the program plan, because it requires either facility modification or extended test-article duration to accumulate equivalent hours at partial conditions. This is the schedule driver.

The xenon feed system reads at TRL 8 and requires no advancement. It has flown on related missions and requires only tuning for the specific propellant loadout and feed pressures.

The system integration element reads at TRL 4 — the weakest link. Component-level maturity does not translate to system-level maturity without integrated test, and no full-power ground test facility is currently available that can host the complete propulsion module operating at the thermal and power conditions representative of flight. The 4→7 transition is the program’s most consequential uncertainty, because it requires either qualification of a new facility, scheduling of an existing facility that is externally controlled, or acceptance of residual risk through analysis-based qualification — a path that reviewers have historically treated with skepticism.

The maturity profile, read together, produces a recommendation that the arithmetic average would have suppressed. The weakest link is the system-integration element at TRL 4; the schedule driver is the PPU’s three-level gap; the thruster is the simplest of the gaps to close. The program’s leading investment in risk retirement should therefore go to PPU thermal qualification and to resolving facility access for integrated-system testing, in that order. The thruster, which a spec-level review would have flagged first as the highest-profile component, is actually the least urgent of the three incomplete elements because its advancement path is well understood and its facility access is relatively secured.

This is the non-obvious insight a rigorous TRL assessment produces: the component that dominates the program narrative is rarely the component that dominates the readiness timeline. An executive who sees the system-integration TRL 4 clearly, rather than averaged out of view, can allocate attention and budget correctly.

Where It Earns Its Keep and Where It Falls Short

The method’s strength is legibility. A rigorous TRL assessment converts a murky “how mature is this?” question into a structured answer that survives challenge and supports decisions across procurement reviews, investment committees, and regulatory assessments. Its global adoption means that findings travel well across organizational boundaries — an ESA-read TRL 6 means substantially what a NASA-read TRL 6 means, after harmonization, and the common vocabulary lets heterogeneous teams converge on shared understanding.

Its weaknesses are consistent with its disciplinary limits. TRL is a maturity metric, not a suitability metric. A technology at TRL 9 is ready to fly; it is not necessarily the right technology to fly. Pairing with technical benchmark comparison is how fitness-for-purpose questions are answered properly. The scale does not capture manufacturing readiness, which is addressed by the parallel Manufacturing Readiness Level (MRL) framework; a TRL 9 technology with MRL 4 cannot be produced at scale regardless of how flight-proven it is. It does not capture cost-effectiveness, market demand, or regulatory acceptability, each of which belongs to its own method.

The scale maps awkwardly onto software-dominant or algorithmic innovations. The hardware-centric TRL definitions — prototype in relevant environment, system in operational environment — do not translate cleanly to continuously deployed software or to machine-learning systems whose performance is a function of training data rather than of physical demonstration. Practitioners working in those domains use either adapted scales (NASA has published software TRL variants) or pair TRL with method-specific maturity frameworks. False precision is a recurring hazard; the difference between adjacent TRLs is sometimes subjective, and rigor in evidence grading is what keeps the numbers honest.

The method also does not account for technology regression — the loss of a qualified supply chain, the obsolescence of a test facility, the retirement of specialized workforce. A technology can move backward on the scale even without active decisions, and assessments that do not track regression risk miss a real category of program exposure.

The library treats TRL as a hub. Its scores feed the data-confidence column in technical benchmark comparisons. Low-TRL elements generate risk-matrix entries with likelihood calibrated by gap magnitude. Gap-closure cost estimates feed investment analyses. S-curve lifecycle analysis provides the trajectory overlay that TRL’s static read does not. A TRL assessment consumed in isolation from these neighbors answers “how mature?” well but leaves “what should we do about it?” underdetermined.

For the Practitioner