PiLogic partners with Air Force lab to test satellite fault-prediction software

2026-06-16 · Source: SpaceNews · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems · Depth: Intermediate, short

Summary

PiLogic, a startup specializing in artificial intelligence software for satellite fault prediction, has partnered with the U.S. Air Force Research Laboratory (AFRL) under a two-year Cooperative Research and Development Agreement (CRADA). This collaboration will test PiLogic's technology on an AFRL CubeSat experiment launched in 2022, focusing on spacecraft electrical and power systems. PiLogic's software, described as "exact AI," uses probabilistic reasoning and automated causal analysis to detect anomalies, predict failures, and recommend corrective actions. This approach contrasts with traditional rules-based systems by determining underlying causes from combined engineering models and probability theory, offering explainable understanding. The partnership aims to advance spacecraft autonomy and health monitoring, addressing the aerospace and defense sector's need for predictable AI systems that operate effectively with limited datasets, unlike large language models. The company raised \$4 million in seed financing in 2025 and is pursuing opportunities with NASA and the U.S. Space Force.

Key takeaway

For AI Engineers or Research Scientists developing critical systems with limited data, PiLogic's "exact AI" approach offers a compelling alternative to traditional machine learning. You should investigate probabilistic reasoning and automated causal analysis to build more predictable and explainable fault detection systems. This method is particularly relevant for aerospace, defense, or industrial applications where AI "hallucinations" or unpredictable behavior are unacceptable risks. Consider piloting such solutions for aging infrastructure.

Key insights

PiLogic's "exact AI" uses probabilistic reasoning for explainable satellite fault prediction, addressing data scarcity and predictability needs in aerospace.

Principles

• Probabilistic reasoning enhances AI predictability.
• Causal analysis improves fault diagnosis.
• Explainable AI is critical for defense applications.

Method

PiLogic's software analyzes onboard sensor data, combining engineering models, physics-based relationships, and probability theory to determine the most likely underlying cause of spacecraft anomalies and predict failures.

In practice

• Evaluate "exact AI" for critical systems.
• Apply causal analysis to complex diagnostics.
• Use probabilistic AI where data is limited.

Topics

• Satellite Health Monitoring
• AI Fault Prediction
• Probabilistic AI
• Causal Analysis
• Spacecraft Autonomy
• AFRL Partnership

Best for: AI Scientist, AI Engineer, Research Scientist, Director of AI/ML

Related on AIssential

• Relational Structural Causal Models — Relational Structural Causal Models enable AI to perform causal and combinatorial reasoning in dynamic, multi-object environments.
• We need a ‘Planetary Neural Network’ for AI-enabled space infrastructure protection — A Planetary Neural Network (PNN) using AI is crucial for managing escalating space debris and security threats.
• Navigating Common Pitfalls in Data Science: Lessons from Pierpaolo Hipolito - ML 183 — Understanding data paradoxes and causal reasoning is crucial for building accurate and interpretable machine learning models.
• A Causal Argumentation Method for Explainability of Machine Learning Models — Integrating causal discovery with argumentation frameworks provides "why" explanations for ML model predictions.
• On-board Telemetry Monitoring in Autonomous Satellites: Challenges and Opportunities — A framework uses "peepholes" from neural networks to provide explainable fault detection in autonomous satellite telemetry.
• Enhancing AI and Dynamical Subseasonal Forecasts with Probabilistic Bias Correction — Probabilistic Bias Correction (PBC) significantly improves subseasonal weather forecast accuracy and extreme event prediction.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by SpaceNews.