Living Failure Knowledge Engine
Continuously structures failure evidence into reusable reliability knowledge.
Built for Engineers and Asset Owners
Turn engineering failure knowledge into actionable reliability intelligence.
Risk on Radar turns fragmented engineering evidence into adaptive reliability intelligence for FMEA, root-cause analysis, and system-level risk assessment.
System-Level Risk Analysis
Models subsystem dependencies, propagation paths, and cascading risk.
Cross-Domain Failure Intelligence
Transfers failure patterns across operating contexts and engineering domains.
0+
Papers indexed
0+
Failure mechanisms
0+
Component types
0+
Engineering systems
From evidence to decision support
How the intelligence layers become engineering decisions
Sources
Evidence sources
Papers, standards, industrial reports, sensor data, and NDT records.
Structure
Structured knowledge
Failure modes, causes, effects, controls, and operating context.
Reason
Reliability reasoning
Subsystem dependencies, similarity matching, and cross-domain transfer.
Decide
Engineering decisions
FMEA, RCA, predictive maintenance, and operational risk assessment.
FMEA is only as good as the team in the room.
The methodology is sound. The execution breaks when failure knowledge depends on memory, old templates, and whoever happens to be available.
1
Search by component or system
Enter a component, operating context, or system.
2
Review failure modes with citations
Review ranked failure modes with source-linked evidence.
3
Build your FMEA row by row
Accept, edit, or reject suggestions and export traceable rows.
01
Schedule Workshop
3–6 weeks of expert coordination per analysis
02
Copy Old FMEA
No fresh risk analysis: stale knowledge locks in old gaps
03
Fill from Memory
S/O/D scores are entirely subjective, no external evidence
04
File & Forget
Static document, never updated, knowledge never reused
Incomplete failure modes
FMEA may not comprehensively capture all failure modes. Used as a top-down tool it surfaces only major risks, leaving critical gaps undocumented.
Engineering Failure Analysis · Elsevier, 2011Subjective, unreproducible scoring
Human bias dominates S/O/D scoring in both experts and novices. Different teams produce different RPN values for the same risk, giving a false impression of objectivity.
Engineering Failure Analysis · Elsevier, 2012Knowledge never transfers
Acquiring failure modes is labour-intensive and error-prone due to incomplete data and lack of standard vocabulary. Lessons from one project rarely reach the next.
Arabian-Hoseynabadi et al. · Int. J. Electrical Power & Energy Systems, 2010Evidence-backed FMEA
From blank row to traceable failure intelligence in minutes.
bearing
258 papers · 5 failure modes
Failure mode
Severity
Evidence
Rolling element fatigue
CRITICAL
104 papers
Lubrication breakdown
HIGH
62 papers
Fretting corrosion
HIGH
41 papers
Cage fracture
MEDIUM
29 papers
Brinelling
MEDIUM
22 papers
1
Search by component or system
Enter a component, operating environment, or system. Risk on Radar searches 2,800+ indexed papers instantly.
2
Review failure modes with citations
A ranked list of failure modes, causes, effects, and controls, each linked to its source. Fully traceable.
3
Build your FMEA row by row
Accept, edit, or reject each suggestion. The engineer owns every decision. Export or work directly in the platform.
Cross-domain transfer
Failure knowledge should move across industries, not stay trapped inside them.
Risk on Radar adapts source-domain failure evidence to target operating contexts, so patterns found in one engineering system can inform risk assessment in another.
Source evidence
Wind turbine drivetrain
Industrial rotating machinery
Rail traction systems
Transferable signature
Bearing fatigue under cyclic load
Failure mode
Cause pattern
Evidence confidence
Target context
Turbofan accessory gearbox
Marine propulsion assets
High-speed manufacturing lines
Load spectrum
Temperature
Lubrication regime
Duty cycle
Environment
Maintenance history
Why reliability software needs living evidence.
Static libraries
Fixed failure libraries become outdated quickly and rarely absorb new literature, standards, field reports, or operating evidence.
Siloed knowledge
Failure evidence stays scattered across papers, reports, teams, standards, and vendor tools instead of becoming reusable structured knowledge.
No context transfer
Conventional tools rarely reason about whether a failure pattern from one asset, domain, or duty cycle applies to another operating context.
Weak traceability
Suggested failure modes, causes, and controls often lack clear provenance, making it difficult to defend or update reliability decisions.
Case Study
Turbofan Engine
assembled
Failure mode evidence across the Turbofan engine
Each sector is an engine component; the radial stack shows how many peer-reviewed papers report each failure mode for that component, drawn from a structured review of the turbofan failure-analysis literature.
From blank row to evidence-backed FMEA in minutes.
1
Search by component or system
Enter a component, operating environment, or system. Risk on Radar searches 2,800+ indexed papers instantly.
2
Review failure modes with citations
A ranked list of failure modes, causes, effects, and controls, each linked to its source. Fully traceable.
3
Build your FMEA row by row
Accept, edit, or reject each suggestion. The engineer owns every decision. Export or work directly in the platform.
bearing
1,508 papers · 5 failure modes
Failure mode
Severity
Evidence
Rolling element fatigue
CRITICAL
847 papers
Lubrication breakdown
HIGH
312 papers
Fretting corrosion
HIGH
188 papers
Cage fracture
MEDIUM
97 papers
Brinelling
MEDIUM
64 papers
Common questions.
Risk on Radar indexes peer-reviewed failure literature into a structured database of failure modes, causes, effects, and controls, then surfaces that knowledge inside your FMEA workflow. Search by component or system and get evidence-backed suggestions you can accept, edit, or reject. It's a copilot, not an autopilot.
Most tools (APIS IQ, Relyence, ReliaSoft) focus on structured authoring and standards compliance, but they start from a blank row. Risk on Radar's differentiator is the external intelligence layer: a continuously updated knowledge graph from failure literature that no incumbent currently owns.
No. Every suggestion is reviewed and approved by the engineer. The platform surfaces documented evidence so engineers can make better-informed decisions. Full source traceability is maintained for every row.
The initial focus is mechanical and industrial systems: rotating equipment, fluid systems, structural components, and process plant. Standards support includes AIAG-VDA (Action Priority scoring), ISO 26262, and IEC 61508. Automotive, aerospace, medical, and energy verticals are in scope.
Via structured ingestion of open-access and licensed journal literature using publisher TDM APIs (Crossref, Elsevier, Springer Nature). Each record is normalized into a component → failure mode → cause → effect → control taxonomy with source DOI, confidence scoring, and evidence text spans. Human-in-the-loop validation is applied throughout.
Risk on Radar draws from failure evidence across more than 100 scientific journals covering failure analysis, sensors and measurement, detection technologies, NDT, and reliability methods for controls and corrective actions.
Machine-learning models compare component context, operating conditions, failure mechanisms, and evidence signatures to identify similar patterns across domains.
Three phases, one mission.
Phase 1
In progress
Failure Intelligence Engine
A living knowledge graph indexed from 2,800+ peer-reviewed failure papers across 75+ mechanical component types. The database that makes evidence-backed FMEA possible.
- Structured ingestion via Crossref, Elsevier TDM, and Springer Nature APIs
- Component → failure mode → cause → effect → control taxonomy
- DOI-linked citations with confidence scoring and evidence text spans
- Human-in-the-loop validation at every ingestion stage
Phase 2
Planned
System-Level Risk Analysis
Cross-system failure propagation and dependency mapping. Understand how a single component failure cascades through the full system, before it happens in the field.
- Graph-based failure propagation modelling
- Cross-component dependency visualisation
- Interface failure mode library
- Utilized for existing and user-defined engineering systems
Phase 3
Planned
Cross-Domain Failure Intelligence
Unified failure knowledge across automotive, aerospace, industrial, and energy. What failed in a wind turbine bearing might prevent a failure in a jet engine. We connect the dots.
- Multi-domain taxonomy alignment
- Cross-industry failure pattern detection
- Domain-adapted severity and occurrence tables
- Standards mapping: ISO 26262, IEC 61508, DO-178C
rısk on radar.
Get early access.
Join a small group of reliability and quality engineering teams shaping the product.
- Priority platform access
- Input on the product roadmap
- Founding team pricing
- Dedicated onboarding support
You're on the waitlist.
We'll reach out as soon as early access opens.