Iec 62304 Software Risk

IEC 62304 Software Risk Management Process

Purpose

Guide implementation of the software-specific risk management process (IEC 62304:2006+A1:2015, Clause 7) alongside ISO 14971:2019—identifying how software items contribute to hazardous situations, defining and verifying software risk controls, and re-analysing risk when software changes.

When to Apply

• Architecture or requirements work for Class B/C software items.
• Introducing or updating SOUP, interfaces, or control logic that mitigates hazards.
• Any software change that may add causes, weaken controls, or alter hazardous situations (7.4).
• Linking hazard analysis outputs to software requirements and tests.

Requirements (testable)

1. Hazardous-Situation Mapping [Class B, C]: For each hazardous situation from device risk analysis, identify software items that could contribute (direct failure or failure of a software-implemented control). Rationale: scoped software risk analysis per 7.1.1.
2. Potential Causes [Class B, C]: Document plausible causes per item—including specification gaps, implementation defects, third-party library (SOUP) behaviour, hardware interactions, and foreseeable misuse. Rationale: 7.1.2.
3. SOUP Anomaly Review [Class B, C]: When SOUP failure is a potential cause, review supplier-published anomaly/issue lists for the version in use and assess whether known issues could lead to a hazardous situation. Rationale: 7.1.3.
4. Risk File Records [Class B, C]: Record identified causes and related items in the risk management file (ISO 14971). Rationale: 7.1.4.
5. Control Definition [Class B, C]: For each documented cause, define risk control measures per ISO 14971; note whether implemented in software, hardware, procedures, or labelling. Rationale: 7.2.1.
6. Software-Implemented Controls [Class B, C]: When a control is implemented in software—add to software requirements, assign safety class to the implementing item based on the risk the control addresses, and develop per Clause 5. Rationale: 7.2.2.
7. Control Verification [Class B, C]: Verify each software-related control and document results; review whether the control could introduce a new hazardous situation. Rationale: 7.3.1.
8. Software Hazard Traceability [Class B, C]: Maintain trace links: hazardous situation → software item → cause → control → verification evidence. Rationale: 7.3.3.
9. Change Impact — New Causes [Class A, B, C]: On software or SOUP change, analyse whether new potential causes or hazardous contributions are introduced. Rationale: 7.4.1.
10. Change Impact — Existing Controls [Class B, C]: Analyse whether the change could interfere with existing software risk controls. Rationale: 7.4.2.
11. Re-run Risk Activities [Class B, C]: After change analysis, repeat relevant 7.1–7.3 activities when analysis shows new or weakened risk. Rationale: 7.4.3.

Recommended Practices

• Keep risk-register.yaml entries keyed by HZ-###, RISK-CTRL-###, and software item names.
• Pair each software control with at least one negative-path test.
• Reconcile SOUP CVE/advisory feeds with the formal anomaly list review.
• For controls that only run in software, avoid assigning a lower class than the risk they address.

Patterns

Software cause and control (YAML):

hazardous_situation: HZ-07
software_item: pump_control
potential_causes:
  - id: CAUSE-07a
    type: implementation_defect
    desc: rate limit not applied on low battery path
  - id: CAUSE-07b
    type: soup
    soup: tiny-rtos@1.4.2
    desc: timer drift under load (supplier anomaly TR-12 reviewed 2026-05-01)
risk_controls:
  - id: RISK-CTRL-19
    location: software
    requirement: REQ-62304-102
    verification: [TEST-410, TEST-411]

Control verification note:

RISK-CTRL-19 verification: TEST-410 PASS, TEST-411 PASS
New hazardous situation review: none identified

Anti-Patterns (risks)

• Treating ISO 14971 file as complete without Clause 7 software cause analysis -> risk: missing software-specific causes.
• Skipping SOUP supplier anomaly review -> risk: known defects unmitigated.
• Lowering item class below the risk addressed by a software control -> risk: under-verification of critical mitigations.
• Change merges without 7.4 analysis -> risk: silent control breakage.

Verification Checklist

• [ ] Software items mapped to hazardous situations for Class B/C scope.
• [ ] Potential causes documented per item; SOUP causes include supplier anomaly review.
• [ ] Controls defined; software controls reflected in requirements with appropriate class.
• [ ] Each software control has documented verification; new hazardous situations considered.
• [ ] Trace chain recorded: situation → item → cause → control → verification.
• [ ] Software/SOUP changes include 7.4.1 analysis; B/C changes include 7.4.2–7.4.3 when applicable.

Traceability

• Use HZ-###, CAUSE-###, RISK-CTRL-###, REQ-###, TEST-### consistently with REG-ISO14971 and DOC-TRACEABILITY.
• Store Clause 7 records in the risk management file or linked artifacts under configuration control.

References

• IEC 62304:2006+A1:2015, Clause 7 (software risk management process).
• ISO 14971:2019 (device risk management; prerequisite skill REG-ISO14971).
• IEC TR 80002-1:2019 (application of risk management to medical device software; informative).
• REG-IEC62304 (lifecycle hub), ARCH-SAFETY-CLASS (classification), DOC-SW-REQ (5.2 risk controls in requirements).

Changelog

• 1.0.0 (2026-05-21): Initial skill for IEC 62304 Clause 7 software risk management process.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: AminAlam
• Source: AminAlam/meddev-agent-skills
• License: MIT

Install and usage instructions live in the source repository linked above.