Predicted → Tested → Flown: linking simulation, qualification and flight telemetry
In short
Predicted → Tested → Flown follows one requirement through its whole evidence life: the model that predicted the behaviour, the ground test that measured it, and the flight telemetry that shows how it actually behaved.
Linking the three lets engineers see margins, correlate models and feed real performance into the next design. In AEAIP, digital twins and telemetry are evidence sources, not separate platforms.
01
Three kinds of evidence for one requirement
Predicted
Analysis, simulation, digital twins
Strength · Covers cases that cannot be tested: full orbits, worst-case combinations, thousands of dispersions.
Limit · Only as credible as the model, its validation range and its inputs.
Tested
Qualification and acceptance testing on the ground
Strength · Measured behaviour of real hardware under controlled, often enveloping, conditions.
Limit · A finite set of conditions, on a specific article and configuration.
Flown
Telemetry, health monitoring, anomaly records
Strength · The actual behaviour of the actual system in the actual environment.
Limit · Arrives after the decisions it would have informed, and only for conditions the mission meets.
- Requirement
- Simulation / Digital Twin Evidence
- Ground Qualification Evidence
- Acceptance Evidence
- Flight Telemetry
- Actual Operational Performance
02
Worked example: a battery temperature requirement
Requirement
THM-008
Battery 5 °C to 30 °C in all operating modes
Requirement
Predicted
Thermal model, hot case
27.5 °C predicted
Simulation / digital twin
Tested
Qualification TVAC, hot plateau
28.9 °C measured; model correlated (+1.4 °C)
Ground qualification
Accepted
Flight unit acceptance TVAC
28.4 °C measured
Acceptance
Flown
Flight telemetry, first 90 days
21 °C to 26 °C observed
Operational evidence
Fictitious values. Operational evidence is a long-term direction (Concept); no telemetry integration exists today.
The thermal model predicted a hot-case battery temperature of 27.5 °C against a 30 °C limit. Qualification thermal-vacuum testing measured 28.9 °C, so the model was correlated: it had been running about 1.4 °C cool. The corrected prediction still met the requirement, with less margin than first thought.
Acceptance testing of the flight unit measured 28.4 °C. In the first ninety days of flight, telemetry showed 21 °C to 26 °C, because real orbits and attitudes rarely sit at the worst case.
Linked together, these records answer questions none of them answers alone: how much margin the mission really has, whether the correlated model is trustworthy for the next design, and whether the hot case should be revisited if the operations profile changes.
03
Model correlation closes the loop between predicted and tested
Models are not assumed correct. Thermal models are correlated against thermal-vacuum measurements; structural models are updated against modal survey and vibration results. Correlation changes the model, which changes every prediction made with it.
That makes the model version part of the evidence's configuration. A prediction from the uncorrelated model and one from the correlated model are different evidence, even for the same requirement. This is configuration-aware validity applied to simulation; see configuration-aware evidence validity.
04
Digital twins are evidence sources
A digital twin is a model kept in step with a specific physical system using that system's data. It is useful exactly to the degree that its outputs can be trusted, which depends on:
- its version, and the data it was last updated with;
- the range over which it has been validated against test or flight data;
- the lineage of its inputs, including telemetry provenance.
AEAIP treats a digital twin as one more producer of evidence, held to the same provenance and configuration rules as a test bench. It is not positioned as a separate GAMYAA platform.
05
Telemetry as operational evidence
After launch, commissioning activities verify functions that could not be fully verified on the ground, such as deployments and end-to-end payload performance. Through the mission, telemetry trending shows margins and degradation, and anomaly records show where the system behaved unexpectedly.
When an in-flight anomaly occurs, the investigation goes back to the ground evidence: what was predicted, what was tested, on which configuration. A connected record shortens that investigation and makes its lessons reusable for the next mission's requirements and margins.
References / Engineering Sources
Explanations on this page are original GAMYAA paraphrase. Standards are referenced, not reproduced; paid standards link to their publisher. Referencing a framework does not imply compliance with it.
- 01NASA Systems Engineering Handbook — 5.3 Product Verification (opens in a new tab)
NASA
Verification methods, qualification, acceptance and recording the product and requirement versions verified.
- 02NASA Systems Engineering Handbook — 5.4 Product Validation (opens in a new tab)
NASA
Validation against stakeholder expectations in the intended environment.
- 03GSFC-STD-7000B, General Environmental Verification Standard (GEVS) (opens in a new tab)
NASA Goddard Space Flight Center
Environmental verification programmes and test levels for GSFC payloads, subsystems and components.
- 04ECSS-E-ST-10-03C, Space engineering — Testing (opens in a new tab)
European Cooperation for Space Standardization
Verification by testing of space segment elements and equipment before launch.
- 05ECSS-E-ST-10-02C Rev.1, Space engineering — Verification (opens in a new tab)
European Cooperation for Space Standardization
Verification methods (test, analysis, review-of-design, inspection), levels and stages.
- Published by
- Gamyaa AeroSystems — Engineering Assurance Research
- Last reviewed
- Status
- Status: Concept