Panasonic Avionics Accelerates Aircraft Diagnostics with Agentic AI on AWS
AWS ML Blog · rss · 2026-08-22
Panasonic Avionics Corporation partnered with AWS and the AWS Generative AI Innovation Center to build an agentic AI system that significantly reduces diagnosis time for In-Flight Entertainment and Connectivity (IFEC) systems. The solution cuts analysis cycles from hours to minutes while maintaining high accuracy.
Challenge:
- Diagnosing issues across thousands of unique deployment configurations required manual correlation of logs, metrics, and ticketing data, demanding deep institutional knowledge and leading to long Mean Time to Resolve (MTTR).
Multi-Agent Architecture:
The system processes data through three distinct layers:
- Trend Analyzer: Continuously evaluates KPIs and service degradation metrics. It uses fleet-wide relationship modeling to proactively identify emerging issues, such as gradual degradation in specific configuration variants.
- Parallel Diagnostic Agents: Triggered by the Trend Analyzer, these agents investigate simultaneously:
- Correlation Analyzer: Detects recurring patterns across deployments with shared configurations.
- System Checks: Validates metadata and service status against ticketing workflows.
- Log Analyzer: Matches current log patterns against a library of known failure modes.
- Summarizer: Powered by Anthropic Claude on Amazon Bedrock, this agent synthesizes findings into structured diagnostic reports, including root cause hypotheses, impact analysis, and prioritized remediation recommendations.
Technical Implementation:
- Orchestration: Amazon SageMaker uses LangGraph (an open-source framework for stateful AI workflows) to orchestrate agents.
- Agent Framework: The Strands Agents SDK (open-source Python framework) provides agent implementation and execution capabilities.
- Context & Memory: A domain ontology normalizes terminology across fleet variants. Semantic search on Amazon RDS with pgvector retrieves similar past incidents and resolutions, providing the system with institutional memory.
- Data Pipeline: AWS Glue and Amazon EMR handle the ETL pipeline into an Amazon S3 data lakehouse using Apache Iceberg.
More from coding & agent
- AI Automation Architect Prompt: Design Workflows and Tool Selection — HeyZoyaKhan · 2026-08-22
- Developer criticizes closed-source models for hiding Chain of Thought — johnnyApplePRNG · 2026-08-22
- An Agent is 3 Layers: Business Logic, Harness, and Infrastructure — blaizedsouza · 2026-08-22
- Every Agent gains computer control, experiments reveal high token costs — every · 2026-08-22
- Grok Build 1.0.8 ships: concurrent subagents start faster, no more frozen sessions — kevinnbass · 2026-08-22
- GitHub Hit: A Massive Collection of Claude Custom Skills and Resources — tom_doerr · 2026-08-22