Skip to content

ORR fatigue risk assessment tools

Fatigue risk assessment tools — including roster-based scoring and biomathematical models — can support structured review of shift patterns. ORR appendix material on benefits and limitations explains what such tools can and cannot do in rail fatigue management.

This page summarises ORR’s view. It does not promote software, claim ORR or HSE approval for any tool, or state mandatory FI/RI thresholds.

What fatigue assessment tools can help with

Section titled “What fatigue assessment tools can help with”

ORR guidance identifies possible benefits of fatigue risk assessment tools and the biomathematical models that may underpin them:

  • Assessing likely fatigue from a current working pattern — to help decide whether further reduction measures may be reasonably practicable
  • Comparing roster options when changes are proposed — timetable changes, new services, or maintenance regimes
  • Supporting consistency — documented, repeatable assessment alongside competent judgement
  • Communicating exposure — helping planners and managers discuss relative fatigue likelihood between options

Tools can be decision-support within a wider fatigue risk assessment and FRMS. They do not replace those frameworks.

Many rail organisations use methods derived from HSE RR446 research — often discussed as a fatigue risk index (FRI) with Fatigue Index (FI) and Risk Index (RI) outputs.

See:

Model outputs are estimates based on population averages and input quality. They do not measure a specific person’s alertness at a given moment.

Tools can help organisations:

  • Record assessment rationale when rosters are designed or changed
  • Compare options on a consistent basis
  • Retain evidence for fatigue records and audit trails
  • Support review when planned vs actual operations diverge

Documentation supports governance — it is not proof that work is safe or compliant.

ORR guidance stresses that tools have known limitations, including:

  • Outputs depend on input accuracy — assumed vs actual times, travel, and breaks
  • Models reflect average responses — individual sleep, health, and lifestyle vary
  • Tools may not capture all operational factors — workload spikes, environmental stress, or disruption
  • Low scores do not guarantee alertness; high scores indicate elevated likelihood, not certainty of harm
  • Tools do not replace competent person judgement on whether work should proceed

See limitations of fatigue models for a wider explanation.

HSE has stated it does not specify mandatory FI thresholds for employers. Organisational or contractual thresholds may exist in specific contexts — those should be assessed against primary sources, not presented as universal HSE law.

ORR expects dutyholders to use tools alongside — not instead of:

  • Operational knowledge of the work and workforce
  • Qualitative hazard identification
  • Worker consultation where reasonably practicable
  • Management accountability when elevated risk is accepted

A tool output should inform review, not trigger automatic approval or prohibition without context.

Relationship to FRI, FI/RI and biomathematical models

Section titled “Relationship to FRI, FI/RI and biomathematical models”
Concept Role
Biomathematical model Mathematical representation of fatigue processes (timing, sleep, workload)
FRI method Practical roster-scoring approach informed by RR446 research
FI Modelled probability-style index for high fatigue on a shift
RI Relative comparison against a reference schedule

The public HSE FRI calculator was withdrawn from the HSE website in 2021. Organisations may still use RR446-informed methods with appropriate governance — see HSE FRI calculator withdrawal.

Fatigue assessment tools should not be described as:

  • ORR-approved, HSE-approved, or Network Rail-approved products
  • Proof that a roster is safe or legally compliant
  • Substitutes for fatigue risk assessment under MHSWR
  • Automatic authorisation to proceed with elevated fatigue exposure

Software that scores rosters is one possible implementation — this site does not promote specific products.