TX11 — Software, Modeling, Simulation, and Information Systems¶
Created: 2026-04-05 (session 8) | Updated: 2026-04-05 (session 9 — AI/ML TRL 5-6 analysis added)
Overview¶
TX11 (Software, Modeling, Simulation, and Information Systems) is the second largest technology area in completed SBIR/STTR projects at TRL 7: 48 projects (12.1% of all 397 TRL 7 SBIR completions). Smaller than TX08 (74 projects) but significantly larger than the next group (TX01: 41, TX06: 36).
Data query: portfolio_aggregate(filter={"program":"SBIR/STTR","status":"Completed","trlCurrent":7}, group_by=primaryTx) → confirmed 48 TX11 projects. All 48 retrieved via find_projects(technology_area="TX11", trl_min=7, trl_max=7, status=Completed, program=SBIR/STTR).
Key finding: No explicit AI/ML or machine learning projects appear in this TRL 7 cohort despite TX11 nominally covering intelligent data processing. The cohort is dominated by simulation tools, MBSE platforms, space weather models, and UAS/aviation operations software.
Cohort Breakdown by Theme¶
Theme 1: Space Weather Modeling (5 projects)¶
The largest coherent domain cluster. Multiple vendors building mission-critical space weather analysis tools:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 102510 XB Space Weather Benchmarks | Predictive Science | 2020–2023 | 6→7 | Extreme space weather event probability tool (XB/SIRE2) |
| 125416 XB Extended | Predictive Science | 2022–2023 | 5→7 | Follow-on: couples probabilistic forecasts with event-based forecasting |
| 125299 SIRE2-AC Radiation Climate | Fifth Gait Technologies | 2022–2024 | 3→7 | Space ionizing radiation for arbitrary trajectories (SIRE2 toolkit) |
| 113310 Kamodo Space Weather Models | Ensemble Gov Services | 2021–2024 | 5→7 | Symbolic abstraction layer for containerized space weather models |
| 154620 Rapid Space Weather Data Integration | Meroxa | 2023–2024 | 3→7 | Real-time data pipeline connectors for space weather data |
Predictive Science appears twice — dominant commercial space weather modeling vendor in TechPort. Their XB tool appears to be an operational product with clear stakeholder uptake (National Space Weather Strategy cited in description).
Meroxa (154620): TRL 3→7 in a 6-month Phase I. This is a suspicious TRL jump — Phase I efforts rarely advance from concept demonstration (TRL 3) to prototype demonstration in relevant environment (TRL 7) in 6 months. This is likely a TRL reporting artifact — the final TRL may reflect the underlying Meroxa data pipeline product, not work done in this SBIR period.
Fifth Gait Technologies SIRE2-AC (125299): TRL 3→7 in a 2-year Phase II — more plausible. The SIRE2 toolkit covers radiation environment for any trajectory including crewed vehicles, UAVs, and arbitrary orbits. The ACT-level jump (3→7) over 2 years is fast but plausible if the Phase I established the architecture and Phase II validated against existing models.
Theme 2: UAS / Aviation Operations Software (7 projects)¶
A distinct cluster tied to NASA's UAS Traffic Management (UTM) and urban air mobility research:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 33822 Multi-UAS Command & Control Phase I | OKSI | 2015 | 3→7 | Single-operator commanding of multiple UAS |
| 89985 Multi-UAS Command & Control Phase II | OKSI | 2016–2018 | 5→7 | Phase II maturation |
| 90470 Core Flight Software for UAS | Windhover Labs | 2016 | 4→7 | Flight software architecture (cFS-derivative) for UAS |
| 125308 Chip-Based ADS-B for UAVs | KALSCOTT Engineering | 2022 | 5→7 | FAA-compliant ADS-B In/Out for UAV dense airspace |
| 125609 Long-Term Autonomy for Urban Mobility | Near Earth Autonomy | 2022–2023 | 6→7 | BVLOS autonomous flight in urban dynamic environments |
| 154465 UAS Fire Fighting Data Assimilation | Black Swift Technologies | 2023–2024 | 4→7 | UAS sensor data for wildland fire operations DSS |
| 17852 Razor UAS Test System | Adsys Controls | 2014 | 6→7 | HIL simulation/test system for UAS |
OKSI appears twice (Phase I + Phase II). Near Earth Autonomy is the highest TRL recipient (6→7), developing "last-mile delivery" autonomous flight capabilities under a Phase II-E.
The Black Swift Technologies fire-fighting project (154465) is notable for spanning UAS operations and Earth science applications — wildland fire behavior prediction using airborne sensors in hazardous conditions. TRL 4→7 in a Phase I is again a suspicious jump.
Theme 3: Model-Based Systems Engineering (MBSE) Platforms (6 projects)¶
Multiple vendors building MBSE infrastructure:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 102633 Digital Pipelines Phase I | InterCAX | 2020–2021 | 1→7 | CI/CV/CD framework for complex system models |
| 113175 Digital Pipelines Phase II | InterCAX | 2021–2023 | 4→7 | Distributed digital threads for MBSE |
| 102010 Models as Microservices | InterCAX | 2019–2020 | 1→7 | Web-based model publishing as REST microservices |
| 9356 Analysis & Design Env. | Phoenix Integration | 2012–2014 | 4→7 | ModelCenter + AnalysisLibrary integration framework |
| 102663 EngineeringHub | Phoenix Integration | 2020–2022 | 4→7 | Trusted collaborative MBE sharing platform |
| 113287 Innovative System Modeling | WarpIV Technologies | 2021 | 5→7 | Parallel discrete-event simulation for mission modeling |
InterCAX appears three times and is the dominant MBSE vendor in this cohort. Two of their three projects show TRL 1→7 in Phase I periods (6 months, 2019–2020 and 2020–2021). TRL 1→7 in 6 months is not credible as a single-phase advancement. These TRL values almost certainly reflect the commercial product maturity, not work done in the SBIR period. InterCAX's Digital Pipelines and MaMs are mature commercial products that pre-existed the SBIR grants — the SBIR work adapted them for NASA use.
Phoenix Integration appears twice (ModelCenter ecosystem). Consistent 4→7 TRL over 2-year Phase IIs — more credible trajectory.
Theme 4: Aeroelastic / CFD Analysis Tools (5 projects)¶
Sophisticated numerical simulation tools for aeronautics design:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 8506 Multi-Fidelity MDO Phase I | Mechanical Solutions | 2010 | 3→7 | Coupled fluid-thermal-structural propulsion design |
| 8804 Multi-Fidelity MDO Phase II | Mechanical Solutions | 2011–2013 | 2→7 | Phase II maturation |
| 102241 Aeroservoelastic Modeling FUN3D | ZONA Technology | 2019–2021 | 4→7 | Discrete state-space aeroelastic model from FUN3D CFD |
| 113232 CFD-Based AIC Matrix | ZONA Technology | 2021–2023 | 4→7 | AIC matrix generation for ZAERO commercial software |
| 8609 Multi-Fidelity MDO Env. | ZONA Technology | 2010 | 6→7 | MDO for aerospace conceptual design |
ZONA Technology appears three times — dominant commercial vendor for aeroelastic analysis tools. Their ZAERO commercial software is cited as the "flagship" product in multiple descriptions. The SBIR work integrates NASA's FUN3D (open-source CFD) with ZAERO, extending its validation base. This is a classic SBIR pattern: commercial vendor uses NASA SBIR to integrate with NASA tools, building competitive differentiation.
Mechanical Solutions appears twice (Phase I + II for multi-fidelity propulsion design). Their tool addresses coupled fluid-thermal-structural interactions — relevant to rocket engine development as well as gas turbines.
Theme 5: HPC and Data Compression (4 projects)¶
Computing infrastructure improvements:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 9663 GPU Sparse Matrix Solvers | EM Photonics | 2011–2013 | 3→7 | CUDA-accelerated linear algebra for CFD/climate |
| 17789 Communication-Intensive Compression | Accelogic | 2014–2017 | 3→7 | Novel data compression to reduce memory bandwidth |
| 90421 Memory-Bandwidth-Limited HPC | Accelogic | 2016–2019 | 4→7 | In-memory compression breaking HPC "memory wall" |
| 103101 Large-Scale Numerical Simulation | WarpIV Technologies | 2020–2021 | 4→7 | Parallel discrete-event simulation (N-body, CFD, etc.) |
Accelogic appears twice — consistent focus on memory-bandwidth-limited HPC. Their approach: aggressive lossless compression of numerical data in flight between CPU and memory, enabling higher effective memory bandwidth. Both projects advance TRL by ~3 over 2-3 years — credible trajectory.
EM Photonics' GPU solver (9663): early CUDA implementation for scientific sparse-matrix systems. Now 12 years old — likely obsolete given vendor GPU library improvements (cuSPARSE, etc.), but represented a genuine TRL advance in 2013.
Theme 6: Mission Planning and Science Tools (5 projects)¶
Software for mission design and operations:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 9916 PM/IDE Planning Domain IDE | Stottler Henke | 2011–2013 | 3→7 | IDE for ANML planning domain models (automated scheduling) |
| 113068 Advanced Science Modeling | Ascending Node Tech. | 2021–2023 | 4→7 | Science objective simulation for mission planning |
| 125692 Observation Planning in Spaceline | Ascending Node Tech. | 2022–2024 | 4→7 | Observation planning tool for Earth/astronomical targets |
| 33880 Marshal: Evolving Task Models | SIFT | 2015–2017 | 5→7 | Mixed-initiative tool for maintaining task models during missions |
| 9452 Rapid Model Fitting | Tech-X Corp | 2012–2014 | 4→7 | Inverse problem solver for remote sensing retrieval |
Ascending Node Technologies appears twice (Spaceline platform). Two tools — science objective simulation (Phase II) and observation planning (Phase II) — suggest they built a coherent mission analysis platform across two SBIR grants. This is a typical SBIR platform-building pattern.
Stottler Henke (PM/IDE): Automated planning domain models using ANML. Stottler Henke is a known vendor for NASA automated scheduling (their ASPEN/CASPER systems have flight heritage). This SBIR advances their planning model IDE.
Theme 7: Software Verification and Fault Management (4 projects)¶
Safety-critical software tooling:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 12856 Semantics-Based Verification | Runtime Verification | 2013 | 5→7 | Modular formal verification for flight-critical C/C++ |
| 16727 SPEEDY Specification Assistant | GrammaTech | 2013 | 2→7 | Specification editing/discovery for formal verification |
| 90192 Fault Management V&V | Qualtech Systems | 2016–2018 | 4→7 | Fault management metrics and verification/validation |
| 9542 Physical Modeling Anomaly Diagnostics | Ridgetop Group | 2011–2013 | 5→7 | Model-driven fault diagnostics for electromechanical actuators |
GrammaTech (16727): TRL 2→7 in a Phase I (6 months). Another suspicious jump. GrammaTech is a well-established formal verification firm — this TRL likely reflects their commercial product maturity, not Phase I work.
Runtime Verification (12856): Phase I only, TRL 5→7. More plausible — they may have been integrating an existing academic tool into flight software verification workflows.
Theme 8: Earth and Applied Science Applications (5 projects)¶
Diverse applications using NASA Earth data:
| Project | Lead | Dates | TRL | Focus |
|---|---|---|---|---|
| 18031 Rice Decision Support System | Applied Geosolutions | 2014–2017 | 5→7 | Global rice production monitoring using NASA Earth data |
| 95713 Flood Disaster Response DSS | Remote Sensing Solutions | 2018–2020 | 4→7 | Interoperable flood response decision support |
| 8994 Software-Defined VLBI Correlator | XCube Communication | 2011 | 4→7 | High-speed VLBI data recording/correlation (geodesy) |
| 9577 Petascale Science Data Discovery | SciberQuest | 2011 | 4→7 | Semantic data discovery for large-scale datasets |
| 90593 CloudTurbine Streaming | Cycronix | 2016–2019 | 4→7 | Cloud-based real-time data streaming (cloud-file duality) |
Applied Geosolutions rice monitoring tool: uses Landsat/MODIS data for global rice production tracking linked to food security and commodity markets. This is an applied Earth observation product — relatively mature (5→7), deployed for real-world decision support.
Theme 9: Remaining (7 projects)¶
Miscellaneous tools that don't fit the above clusters: - Sensor skins for surface pressure (Nanosonic, TX11 classification is likely wrong — this is a hardware sensor) - Aeroservoelastic design env. (ZONA Technology, counted above) - Other simulation infrastructure
Vendor Concentration¶
| Vendor | Projects | Domain |
|---|---|---|
| ZONA Technology | 3 | Aeroelastic/CFD tools (ZAERO) |
| InterCAX | 3 | MBSE / Digital Pipelines |
| Ascending Node Technologies | 2 | Mission planning (Spaceline) |
| Phoenix Integration | 2 | Collaborative engineering (ModelCenter) |
| Accelogic | 2 | HPC data compression |
| OKSI | 2 | UAS command and control |
| Predictive Science | 2 | Space weather benchmarks |
| WarpIV Technologies | 2 | Parallel discrete-event simulation |
8 of 27 vendor names appear more than once, indicating a pattern of platform extension via sequential SBIRs rather than one-off tools.
The Missing AI/ML Finding¶
No explicitly AI/ML or machine learning projects appear in the TX11 TRL 7 cohort. This is notable because:
- TX11.4.2 is labeled "Intelligent Data Understanding" — only 1 project (XCube VLBI correlator, 2011) uses this classification, and it is not ML-based.
- NASA has heavily funded AI/ML since 2018 under multiple programs.
- The cohort spans 2010–2024 — post-2018 projects should show AI/ML if it were reaching TRL 7.
Possible explanations (speculative, unverified): - AI/ML projects reaching TRL 7 may be classified in other TX areas (TX17 for autonomy, TX11.6 for data analytics) - The 48 TRL 7 completions may under-sample AI/ML because flight-software-quality AI/ML validation (required for TRL 7) is harder to achieve in a 2-year SBIR Phase II - AI/ML tools in NASA context may be reaching TRL 5-6 but not yet 7 in the SBIR cohort - The space weather tools (Predictive Science, Ensemble) may use ML internally but don't advertise it in descriptions
Counter-query to test: portfolio_aggregate(filter={"program":"SBIR/STTR","status":"Completed","trlCurrent":6,"primaryTx":"TX11"}) — if this shows AI/ML at TRL 6, the explanation is "AI/ML hasn't cleared the TRL 7 bar yet."
TRL Jump Anomalies¶
Several projects show implausibly large TRL advances in short periods. Pattern: small Phase I efforts (6 months) or early SBIR records show TRL 1-2→7:
| Project | TRL Jump | Duration | Likely Explanation |
|---|---|---|---|
| 102633 Digital Pipelines Phase I | 1→7 | 6 months | Product pre-existed SBIR; TRL reflects commercial product |
| 102010 Models as Microservices | 1→7 | 6 months | Same |
| 16727 SPEEDY Phase I | 2→7 | 6 months | GrammaTech commercial product pre-existed |
| 154620 Meroxa Phase I | 3→7 | 6 months | Meroxa data pipeline product pre-existed |
These are not Phase I breakthroughs — they are commercial products adapted for NASA use. TRL fields may represent the vendor's product maturity at time of SBIR engagement, not TRL gained during the effort. This is consistent with the Closed_Out masking finding: SBIR TRL reporting has significant inconsistency, particularly for software tools.
Key Findings¶
-
No AI/ML at TRL 7 in TX11 SBIR completions. The entire 48-project cohort consists of simulation tools, MBSE platforms, space weather models, and aviation software. This is unexpected given NASA's stated AI/ML investment priorities.
-
Vendor concentration is high. 8 vendors each appear twice or more, with ZONA Technology (aeroelastic tools), InterCAX (MBSE), and Predictive Science (space weather) the most prominent. These are platform vendors using SBIR to extend and integrate commercial products.
-
Space weather software is the strongest coherent cluster — 5 projects from dedicated vendors (Predictive Science, Ensemble Gov Services, ASTRA, Fifth Gait, Meroxa). The Kamodo project (Ensemble) is notable for its open-source community model — a symbolic abstraction layer enabling reuse across research, operations, and education.
-
UAS/aviation software cluster is unexpected in TX11. 7 of 48 projects are UAS or aviation operations software — more than in TX15 (aviation). These reflect NASA's UTM/AAM research investment in software and operations infrastructure rather than vehicle hardware.
-
Software V&V tools are represented — 4 projects on formal verification and fault diagnostics for flight-critical software. This is a coherent need (flight software certification) but a small cluster.
-
TRL jump artifacts are common in this cohort, more so than in hardware-heavy TX08. Software TRL self-reporting appears less disciplined than hardware TRL — phase I "7" results should be treated skeptically.
AI/ML at TRL 5-6: The Valley of Death Is TRL 6→7¶
Added session 9, 2026-04-05
The TRL 5-6 cohort confirms and refines the AI/ML finding. Query: find_projects(technology_area="TX11", trl_min=5, trl_max=6, status=Completed, program=SBIR/STTR) → 170 projects total. Keyword search (machine learning artificial intelligence deep learning neural network) → 59 of 170 explicit AI/ML matches (~35%).
At TRL 5-6, AI/ML IS well-represented. The valley of death is TRL 6→7, not 5→6.
AI/ML Theme Breakdown at TRL 5-6¶
Earth science remote sensing (dominant): Multiple land use image classification tools, deep learning for SAR/InSAR analysis, snow monitoring (Applied Research Team), air quality (182938), 3D atmospheric structure from geostationary data. "Multi-Resolution Deep Learning for Land Use Applications" appears as THREE consecutive projects (112815 Phase I → 112979 Phase II → 125667 Phase III) — Phase I ends TRL 5, Phase II ends TRL 6, Phase III ends TRL 6. Three phases, still at TRL 6. This is the clearest example of the TRL 6 ceiling.
Space weather forecasting (ML variant): "Machine Learning Enabled Thermosphere Advanced by HASDM (META-HASDM)" (102863, TRL 4→5), "Geoelectric Field Forecasting with Machine Learning" (102700, TRL 3→6), "Improved Forecasting of Operational Solar/Geomagnetic Indices" (154514, TRL 5→6). These sit at TRL 5-6 — the same cluster that appears at TRL 7 WITHOUT the ML label. The space weather software at TRL 7 (Predictive Science, Kamodo) uses conventional modeling architectures. ML-specific forecasting tools top out at TRL 6.
Physics surrogate models: "CHEM-ML for Non-Equilibrium Chemistry in Hypersonic Flows" (154688, TRL 3→6) — AI/ML surrogate replacing expensive CFD for hypersonic non-equilibrium chemistry. "Surrogation of High-Fidelity Simulation Software" (154622, TRL 3→5) — mesh-based graph neural networks (GNN) for physics sim surrogates. "Generative Adversarial Networks for NDE/SHM Physics Simulations" (113233, TRL 3→5). All at TRL 5-6, none at TRL 7.
Deep learning hardware for HPSC: "Deep Learning Processing Subsystem (DLPS): A HPSC-Compatible Deep Learning Coprocessor" — Phase I (102623, TRL 2→5) and Phase II (113163, TRL 3→6). This is on-board AI inference hardware for NASA's High Performance Spaceflight Computing (HPSC) processor. Phase II reached TRL 6 — the hardware demonstrator exists, but validation in space environment (TRL 7) hasn't happened. This is notable: a purpose-built DL accelerator for space that stalled at TRL 6.
Spacecraft health monitoring: Lunar FLAPPER (113339, TRL 2→5), MaARS multi-agent anomaly resolution (125655, TRL 3→5), SPARK anomaly knowledgebase (154738, TRL 3→5) — all stuck at TRL 5.
Accelerating Design of ML/AI Experiments (158186): TRL 6→6. Phase II effort that started and ended at TRL 6 — no progress on TRL. Could be TRL reporting that "confirmed" TRL 6 without advancing it.
Why AI/ML Stalls at TRL 6¶
TRL 7 requires "prototype demonstrated in a relevant space environment." For software, this typically means integration with a mission or flight-equivalent system. Three structural reasons AI/ML stalls at TRL 6:
-
Training data dependency: ML models require data from their operational environment. For space, this means flight data that doesn't exist until the mission flies. Circular dependency blocks TRL 7.
-
Validation methodology gap: NASA lacks established methods to "prove" an ML model is reliable for flight software. Traditional V&V (formal methods, exhaustive testing) doesn't apply to neural networks. The software V&V tools in the TRL 7 cohort are classical formal methods — they don't cover ML.
-
Commercial product escape valve: Earth science ML tools (land use, snow monitoring, air quality) have commercial markets. Phase III companies commercialize to industry and don't need NASA TRL 7 certification to generate revenue. The NASA pipeline terminates at TRL 6 as they pivot to commercial.
The DLPS Case Study¶
The HPSC-compatible deep learning coprocessor (102623/113163) is the exception that proves the rule. It's hardware not software — so the TRL 7 bar is a hardware test, not a software validation challenge. Yet it stalled at TRL 6. This suggests there's also a demand-side problem: no mission has formally requested an on-board DL coprocessor at sufficient maturity to justify the TRL 7 integration cost.
Quantitative Summary¶
| TRL Level | TX11 SBIR completions | Explicit AI/ML count | AI/ML fraction |
|---|---|---|---|
| 7 | 48 | 0 | 0% |
| 5-6 | 170 | ~59 | ~35% |
| 83 at TRL 6 specifically | 83 | ~25 estimated | ~30% |
The cliff between TRL 6 and TRL 7 in AI/ML is real and structural.
Open Threads¶
- Kamodo ecosystem: Ensemble Gov Services' Kamodo has open-source ambitions. Is it used in production NASA operations? No Infused_To in TechPort.
- Kamodo ecosystem: Ensemble Gov Services' Kamodo has open-source ambitions. Is it used in production NASA operations? No Infused_To in TechPort.
- Digital Pipelines (InterCAX): Phase II (113175, 2021-2023) adds distributed model sharing with "digital threads." This concept aligns with NASA's Model-Based Systems Engineering Initiative — is there a GCD or program-level project consuming this work?
- Near Earth Autonomy (125609): Urban autonomous flight at TRL 7 — is this feeding into specific AAM programs?
Cross-References¶
- topics/sbir-sttr-high-trl.md — parent analysis; TRL jump anomalies pattern
- programs/sbir-sttr.md — program context; Closed_Out masking
- topics/tx08-sensors-instruments.md — TX08 comparison (largest cohort)