Source tier 1
Austroads 2022, Assessing Fitness to Drive, Section 10.2
Primary source for the visual-field rules.
Learning Library Preview
DRIVE Fields is a free Australian clinician decision-support tool built by Dr Simon Chen, ophthalmologist, to help optometrists and ophthalmologists apply Austroads 2022 driving visual-field criteria more consistently.
DRIVE Fields does not grant, deny, or predict a driver licence. It helps a clinician organise visual-field findings against the Austroads criteria. The clinician remains responsible for the clinical assessment, and the driver licensing authority makes the final licensing decision.
Source tier 1
Primary source for the visual-field rules.
Source tier 2
Workflow and submission context where it does not conflict with Austroads.
Source tier 3
Clinician education and practical interpretation support.
Source tier 4
Background evidence for reliability, device layout, and printout behaviour.
The deterministic rule engine does not use free-form AI judgement to decide whether a visual field meets a threshold. Instead, it asks for structured facts such as licence class, test type, reliability, visual acuity, device pathway, and missed test locations, then applies codified rules with traceable explanations.
The app is intentionally conservative. If a device pathway is uncertain, if reliability data are missing, if the printout does not match a supported grid, or if the pattern sits outside the rules that can be safely codified, the correct result is manual review rather than a confident favourable label.
The calculation is only as good as the source printout and the points entered by the clinician. The original visual-field printout remains the evidence for the clinical record or authority form.
The public manual-grid assessment runs in the browser. It is designed for point marking and report generation without uploading patient identifiers or case details to a DRIVE Fields server. Clinicians should still avoid entering patient identifiers into free-text fields and should keep the original printout within their normal clinical record system.
Manual review is a safety feature. It appears when the app cannot make a source-traceable deterministic assessment from the entered data. Examples include unsupported devices, missing fixation monitoring, unreliable false positives, roving pathway ambiguity, neurological patterns, monocular edge cases, and borderline commercial-driver results.