Key takeaways ▼
Accellera Systems Initiative has approved Functional Safety Language Standard 1.0 for release, addressing a persistent challenge in semiconductor development: moving safety information reliably between tools, engineering teams, and companies. Announced October 21, the standard establishes a common language and data model for capturing, exchanging, and integrating functional safety information throughout electronic systems development. FSL 1.0 is available for download.
The announcement matters because safety analysis increasingly spans an extended supply chain. An IP supplier develops a component, a semiconductor company incorporates it into a chip, and a system manufacturer integrates that chip into a larger product. Each participant needs relevant safety information, including assumptions and limitations, to understand how failures could affect the finished system.
Today, those handoffs can involve spreadsheets, proprietary databases, and incompatible formats. Engineers may need to recreate relationships, translate terminology, or manually reconcile different versions of an analysis. Beyond the time involved, every conversion creates an opportunity for omissions or misunderstandings. A common representation could make those exchanges more consistent and easier to automate.
FSL 1.0 focuses on failure modes, effects, and diagnostic analysis, commonly called FMEDA. This analysis examines potential failures, their consequences, and the mechanisms intended to detect or control them. The language captures failure modes, effects, safety mechanisms, technology information, and safety requirements, alongside mappings that connect safety information with design representations.
Its support for hierarchical FMEDAs is particularly consequential. Component analyses can contribute to subsystem and system analyses, helping information travel alongside the products it describes. For suppliers, this could reduce repeated preparation of customer-specific deliverables. For integrators, it could simplify the work of combining information from multiple sources while preserving traceability.
However, reuse requires engineering judgment. A component’s safety analysis depends on assumptions about its operating environment, implementation, and use. Moving that analysis into a larger system does not automatically establish that those assumptions remain valid. FSL provides a structure for exchanging information; engineers must still assess whether the information applies to their design.
Another feature connects early estimates with later evidence. FSL supports both assumption-based and design-mapped FMEDA approaches, allowing analysis to begin before implementation details are complete and evolve as development progresses. It also supports refining estimated diagnostic coverage using results from verification and fault-injection activities.
That connection could help close a troublesome gap between safety documentation and engineering reality. If verification reveals that a safety mechanism detects fewer faults than expected, the analysis should reflect that finding. More direct exchange between verification and analysis tools could make such updates easier, while helping reviewers understand which conclusions rest on estimates and which have supporting measurements.
FSL uses Tcl-based syntax to support machine processing and integration into automated workflows. It also supports calculation and reporting of safety-related failure rates and metrics associated with ISO 26262 and IEC 61508. Those established standards retain their requirements; adopting FSL alone does not demonstrate compliance or certify a product.
The practical opportunity is better coordination between design, verification, and safety engineering. A shared data model could support consistency checks, identify missing mappings, and make changes easier to trace. Accellera’s working group explicitly targets interoperability across development activities and supply-chain layers, from IP through integrated circuits to systems.
Adoption will determine how much of that opportunity becomes routine practice. Tool vendors need compatible implementations, suppliers need dependable export processes, and customers need confidence that imported information retains its meaning. Successful demonstrations should test round-trip exchange, hierarchical integration, and updates after design changes.
Bottom line: safety decisions depend on evidence that remains understandable as products and organizations change. FSL 1.0 gives the industry a common foundation for carrying that evidence forward. Its value will emerge through reduced translation work, clearer traceability, and stronger connections between documented assumptions and verified behavior, helping engineers spend more time evaluating safety and less time reconstructing data.
Also Read:
Share this post via: