Collaboration Brief: Dr. Manil Maskey & SSAI

Partner Profile: SSAI

Founded in 1977, Science Systems and Applications, Inc. (SSAI) has provided scientific, engineering, and IT support for nearly 50 years. SSAI's expertise spans instrument engineering through operational data systems, making them well-suited to partner with Dr. Maskey's data-centric AI initiatives.

1977Founded

575+Employees

50 YearsNASA/NOAA Support

Tier-1NASA Supplier

Executive Summary

This brief outlines a strategic alignment between Dr. Manil Maskey (NASA MSFC) and SSAI. By combining Dr. Maskey’s leadership in Foundation Models (Prithvi, Surya) and the VEDA platform with SSAI’s mission operations and data systems expertise, the partnership aims to mature AI from research (TRL 4) to operational deployments (TRL 7+).

Capability Synergy

Dr. Manil Maskey (NASA MSFC)

AI Leadership: Lead of SMD AI Team & ODSI; advocate of data-centric approaches.[1]
Foundation Models: Prithvi WxC and Surya, enabling large-scale weather/heliophysics modeling.[2][3]
VEDA: Platform for "compute-to-data" analytics and hosted workspaces.[4]
Standards: Involvement in Croissant metadata and ML-readiness standards.[5]

SSAI Capabilities

Mission Operations: 50-year heritage supporting NASA/NOAA (e.g., SARSAT, VIIRS, OMPS).[6]
Scientific Products: Freshwater Sensing Program (STREAM), Worldview/GIBS visualization systems.[7]
Heliophysics Support: Systems engineering and operational integration; NAIRAS is NASA's model which SSAI integrates where applicable.[8]
End-to-End Engineering: Instrument design, calibration, and data assimilation.[9]

Success Metrics (Deep Dive)

Click a metric card to expand evidence and benchmarks.

Science Impact

RMSE & F1 Improvements

Goal: Use Maskey's Foundation Models to improve SSAI operational baselines.

Evidence:

Hurricane Intensity: Deep learning estimator achieved 7.8 knots RMSE vs observed.[10]
Solar Flare Forecasting: Surya achieved F1 = 0.561 vs ResNet50/AlexNet baselines.[11]

Application: Fine-tune models to reduce error in STREAM and other SSAI products.

Operational Efficiency

Reduced Labeling Time

Goal: Reduce manual annotation costs for SSAI projects.

Evidence:

87% Data Reduction: Prithvi experiments matched benchmarks with 13% of original data.[3]
Cost Implication: Prior large labeling efforts cost millions; foundation models enable few-shot workflows.[14]

Application: Apply to Freshwater Sensing workflows for faster model updates.

Readiness Level

TRL Progression

Goal: Move AI from TRL 4 to TRL 7+ for operational use.

Evidence:

Maskey transitioned a research estimator to a production AWS portal for real-time ingest.[16]
VEDA is a stable cyberinfrastructure hosting high-visibility dashboards, ready for ops deployment.[4]

Application: Validate models in SSAI's operational environment to bridge research → operations.

Joint Roadmap

Phase 1 (Pilot): Fine-tune Prithvi WxC on STREAM water quality data. Target: F1 > 0.8 with 50% less labeled data.
Phase 2 (Integration): Establish SSAI hosted workspace within VEDA to demonstrate reproducible cloud-native workflows.
Phase 3 (Standardization): Adopt Croissant metadata across SSAI archives to ensure ML-readiness.[18]

Collaboration Engine: From Research to Ops

1. Foundation Model Fine-Tuning

The Gap

Prithvi WxC needs downstream validation and tuning for operational hydrology tasks.

SSAI Role

Use STREAM ground-truth data to fine-tune Prithvi and validate performance on water quality and flood mapping.

Outcome

Rapid TRL Advancement for hydrology FMs.

2. Heliophysics: Surya Integration

The Gap

Surya is strong for solar prediction but must be integrated for real-time operational use.

SSAI Role

Integrate Surya outputs with operational pipelines; SSAI supports systems integration with NASA models (e.g., NAIRAS outputs).

Outcome

AI-Driven Space Weather Ops for astronaut & aviation safety.

3. VEDA & Hosted Workspaces

The Gap

VEDA offers compute-to-data but needs operational teams to migrate legacy workflows.

SSAI Role

Migrate pipelines (OMPS, VIIRS) into VEDA and provide hosted workspaces for reproducible cloud-native science.

Outcome

Cloud-Native Production validation of VEDA.

4. ML-Ready Data & Croissant

The Gap

Legacy archives lack machine-learning-ready metadata standards.

SSAI Role

Apply Croissant metadata and automate tagging (e.g., using LLMs) to accelerate discoverability and model training.

Outcome

Discoverable Archives optimized for AI.

SSAI Core Capabilities

Earth Science & Analytics

Support for HLS, Worldview, OMPS pipelines, and the Freshwater Sensing Program (STREAM).

Heliophysics & Space Weather

Integration and operational support for space weather systems; SSAI integrates with NASA outputs as applicable.

Planetary & Lunar Science

Support for Artemis-related projects and planetary modeling.

Engineering & Mission Ops

Instrument design to mission operations, EEE parts, manufacturing, and AI/ML digital twins.