AI Sustainment That Keeps U.S. Combat Aircraft Flying Longer
Lockheed Martin and MANTECH are teaming up to modernize how U.S. combat aircraft are maintained. The focus: real-time monitoring, predictive maintenance, and secure integration across legacy and next-gen platforms. For operations teams, this means higher availability, fewer surprise failures, and clearer logistics planning.
Why this matters for Ops
Aircraft readiness hinges on two things: timely insights and fast decisions. This partnership targets both by pushing analytics to where data is generated and closing the loop with maintainers and supply chains. Expect better mean time between failures, tighter maintenance windows, and fewer grounded assets.
What's being built
The core capability is AI-driven performance monitoring that flags issues before they sideline an aircraft. It ingests onboard telemetry, predicts component degradation, and triggers maintenance actions with parts and labor aligned. Coverage spans older airframes and fifth-gen fighters, improving mission availability without expanding headcount.
"This collaboration between Lockheed Martin and MANTECH will generate a unified team of strengths, capable of creating resilient sustainment ecosystems that can be projected to America and its allies around the world," said Nicholas Smythe, Lockheed's Vice President for Sustainment Business Development.
The stack behind it
Lockheed Martin is bringing its AI Factory-GPU environments, curated data, and MLOps workflows for rapid model deployment. That accelerates model updates, validation, and rollout at scale. MANTECH adds defense-grade systems integration and enterprise modernization, strengthened by its acquisition of Elder Research.
As David Hathaway, MANTECH's Defense Sector President, noted: "This partnership delivers the real-time performance needed to maximize the readiness and operational lifespan of the U.S. combat aircraft fleet."
Security baked in with Zero Trust
Security is not an add-on here. The approach uses Zero Trust principles from the ground up-identity, device, data, and workload checks at every hop. This mirrors current DoD guidance and reduces the risk of inserting AI into mission workflows.
MANTECH's ongoing enterprise cyber work (including a recent $910M task order with U.S. Southern Command) gives the sustainment solution a stable IT backbone-secure software development, application modernization, and cloud migration that won't stall operations.
What changes for sustainment workflows
- From scheduled to condition-based maintenance: actions triggered by actual wear and trend lines, not just flight hours.
- From siloed logs to live data: onboard streams drive alerts, parts ordering, and workforce allocation in near real time.
- From best guess to forecast: models predict failure windows, so ops can plan swaps during low-demand periods.
- From manual checks to automated validation: MLOps ensures models are versioned, tested, and compliant before deployment.
What Ops leaders should track
- Mission capable rate (MC) and fully mission capable rate (FMC)
- Mean time between failures (MTBF) and mean time to repair (MTTR)
- Unscheduled maintenance events per 100 flight hours
- Supply chain cycle time for high-failure components
- False positive/negative rates in predictive alerts
- Patch latency: time from model update to fleet-wide adoption
Implementation checklist (practical and short)
- Inventory your data: map sensors, formats, sampling rates, and gaps by platform.
- Standardize pipelines: set common schemas and quality rules so models generalize across squadrons.
- Stand up MLOps with guardrails: automated tests, drift detection, and rollback plans.
- Embed Zero Trust: enforce identity, device health, and least-privilege access across maintainers and apps.
- Connect supply: link alerts to parts availability and work orders to cut idle time on the ramp.
- Train maintainers: short, scenario-based playbooks for interpreting alerts and closing loops.
Risks to plan for
- Data sparsity on legacy airframes: prioritize high-value components first and build feature stores over time.
- Model drift: set monthly evaluations and threshold-based retraining triggers.
- Alert fatigue: tune thresholds, batch non-critical alerts, and push only actionable recommendations.
- Integration drag: agree on interfaces early-APIs, data rights, and cyber accreditation timelines.
What this signals next
This isn't a lab demo. Lockheed and MANTECH are packaging AI, sustainment, and security into a deployable system that meets defense standards. The expected outcome: more flight hours from existing fleets, fewer surprise breaks, and tighter alignment between maintenance and mission demand.
For operations leaders, the move is clear: get your data house in order, connect alerts to action, and measure what improves availability. The teams that do this well will deliver more sorties with the same aircraft and the same people-just with smarter timing.
If you're building team skills for AI-enabled operations and maintenance, here's a focused catalog of training paths: AI courses by job.
Your membership also unlocks:
