Skip to content

TechPort Outcome Data Quality — Three Failure Modes

Systematic audit of FO technologyOutcomes reveals three distinct ways TechPort outcome data can mislead.

Created: Session 97, 2026-04-07
Last updated: Session 98, 2026-04-07

Summary

Sessions 95-97 conducted a systematic audit of all 30 FO projects with technologyOutcomes records. Three distinct failure modes were identified. These are not minor edge cases — they affect how anyone using TechPort's structured outcome data would understand FO's impact. The bottom line: technologyOutcomes cannot be trusted at face value for FO projects. Every claim requires independent verification.

Failure Mode 1: Aspirational Descriptions

Pattern: Project descriptions claim flights, experiments, or outcomes that never occurred. TRL data contradicts the narrative.

Example: 91344 RAVAN/VACNT (JHU/APL) - Description implies an sRLV flight to mature VACNTs from TRL 3→7 - TRL current = 3 (never advanced) - No independent evidence that an FO sRLV flight occurred - The RAVAN CubeSat mission was funded by NASA ESTO/InVEST, not FO - Correction (Session 96): Downgraded from "Mission Infusion + Technology Heritage" to "FO-Funded but Likely Did Not Fly." Removed $137M from FO pipeline attribution.

Detection method: Compare trlCurrent against trlEnd target. If TRL stayed at start value and description claims a flight, investigate independently.

Prevalence in FO: 1 confirmed case out of 30 projects with outcomes. Possible more in the broader 430-project portfolio.

Failure Mode 2: Aspirational Library Items

Pattern: Library items link to missions or destinations that were intended but where the technology was ultimately descoped or removed.

Example: 12284 JPL Micro Sun Sensor - Library items link to Prox-1 NSSDC page (the intended host mission) - Prox-1 was significantly descoped in 2017 — the sun sensor, propulsion elements, and solar cells were removed before the June 2019 launch - TechPort record creates the impression the sensor flew on Prox-1; it did not - Correction (Session 95): Downgraded from "Mission Infusion (small satellite)" to "Delivered but Descoped."

Detection method: Cross-reference library item mission links against mission manifests and Gunter's Space Page. Descoping is common in small satellite programs and often not reflected in TechPort.

Prevalence in FO: 1 confirmed case. Library items are rare in FO (~17 projects have fileIds), limiting exposure.

Failure Mode 3: Erroneous Outcome Linkages

Pattern: technologyOutcomes contain relatedProjectId fields pointing to completely unrelated projects in different programs, organizations, and technology areas. These appear to be database-level errors, not aspirational claims.

Examples:

91427 AVA (Affordable Vehicle Avionics, ARC)

Outcome Linked Project Actual Subject Related to AVA?
Advanced To 14686 GRC CIF: Carbon nanotube composites Materials science No
Advanced To 15465 MIT STRG: Hierarchical nanocomposites Materials science No
Advanced From 95678 CU Boulder STRG: Autonomous SSB ops Small body navigation No

12184 Heat Pipe Limits (GRC)

Outcome Linked Project Actual Subject Related to Heat Pipes?
Advanced From 146033 JPL CIF: Atmospheric sensing (GPS RO) Radio occultation No
Advanced To 91575 Cornell STRG: Unsupervised ML for rovers Machine learning No

Note: Both "Advanced To" dates on AVA (Feb 2013, Aug 2014) predate the FO project start (Apr 2015), compounding the error.

Detection method: Spot-check relatedProjectId targets. If the linked project's program, organization, and technology area all differ, the linkage is almost certainly erroneous.

Session 98 additions: Deep-investigating the remaining 5 outcome-bearing projects found 4 more erroneous linkages across 4 projects:

14153 Gannon Cosmic-Ray Calorimeter

Outcome Linked Project Actual Subject Related?
Advanced To 14631 ARC CIF: Exoplanet Direct Imaging Starlight suppression optics No
Advanced From 9395 Invocon SBIR: RAID radiation badge Miniature radiation monitor No

91334 LPX Lunar Plant Growth

Outcome Linked Project Actual Subject Related?
Advanced To 14682 GRC CIF: PTFE wire overheating detection Wire diagnostics No
Advanced From 10602 ARC CIF: Lunar Plants Prototype Same concept, same PI YES — LEGITIMATE

12182 UF Telemetric Biological Imaging

Outcome Linked Project Actual Subject Related?
Advanced From 8148 SWaP Optimized Space Transceivers RF transceivers No

106695 UF FLEX Biological Imaging

Outcome Linked Project Actual Subject Related?
Advanced From 88581 EOA Enceladus Organic Analyzer Mass spectrometry No

94005 Northwestern Titanium Nanofoams

Outcome Linked Project Actual Subject Related?
Advanced From 9322 Precision Electrospray Thruster (PETA) Electric propulsion No

Updated prevalence in FO: 6 projects with erroneous linkages out of 30 with outcomes (20%). Only 1 legitimate linkage found in this batch (LPX ← Lunar Plants Prototype). The actual error rate for individual linkage records is higher — most projects have 2+ erroneous records each.

Additional Finding: Temporal Confusion in "Advanced To" Dates

Many "Advanced To" outcomes have dates that predate the FO project start. This means they record the technology's broader history, not the project's downstream outcomes. Of the 6 FO projects with "Advanced To" outcomes:

| Projects with pre-start "Advanced To" | 4 of 6 (67%) | | Projects with mid-project "Advanced To" | 2 of 6 (33%) |

This creates a counting error if anyone tallies "Advanced To" outcomes as evidence of FO downstream impact. Most "Advanced To" records on FO projects reflect upstream lineage misfiled as downstream outcomes.

Implications for KB Methodology

  1. Never count raw technologyOutcomes as impact evidence. Always verify independently.
  2. TRL stagnation (current = begin) is a red flag for aspirational descriptions. Check whether flights actually occurred.
  3. Library item mission links reflect intentions, not outcomes. Cross-reference with mission manifests.
  4. relatedProjectId linkages need spot-checking. At minimum, verify the linked project is in a related field.
  5. "Advanced To" date vs. project start date tells you whether this is upstream history or downstream impact.

These findings are specific to the FO portfolio but likely apply across TechPort. The root causes (aspirational data entry, library items as intended destinations, database linkage errors) are not FO-specific.

Audit Coverage

What was audited Result
All 30 FO projects with technologyOutcomes 3 failure mode types identified
11 canceled zero-TRL-gain projects Normal cancellation records, no RAVAN-type claims
19 completed projects with outcomes (all now audited) 6 projects with erroneous linkages (20%); 1 aspirational description; 1 legitimate lineage found
All "Advanced To" outcomes Most predate project start (temporal confusion); only LPX [91334] ← [10602] confirmed as correct lineage