Transport remains Australia's deadliest workplace. Vehicle incidents accounted for 42% of the 188 worker fatalities recorded in 2024, and Safe Work Australia's figures put human factors, not mechanical failure, behind most serious truck crashes. NTI and NTARC analysis has attributed as much as 63.5% of serious truck crashes to human factors, with fatigue and distraction sitting inside that figure year after year.
Fatigue detection technology watches for the physical signs of micro-sleep and drowsiness inside the cab, and connects to an operator's HVNL fatigue management obligations through the evidence it produces. Crystal Vision video telematics is the evidence layer behind it.
Why fatigue matters for AU heavy fleets
Fatigue is not a soft risk category for Australian heavy vehicle operators. It is a leading contributor to the human-factor crashes that dominate NTARC's Major Incident Investigation Report. It also sits inside the Chain of Responsibility duty every party in the transport chain carries under the Heavy Vehicle National Law.
Operators without fatigue evidence are left defending decisions with memory and paperwork instead of a time-stamped record.
The exposure is not only about the crash itself. NHVR treats telematics that records driving hours, rest breaks, and fatigue events as evidence of executive due diligence. Operators without that evidence are left defending decisions with memory and paperwork instead of a time-stamped record. For fleets running dedicated fatigue and distraction detection, that evidence gap closes. Road-facing and driver-facing AI catches the fatigue and distraction signal before the crash, not after, and every event is logged with a timestamp instead of relying on a driver's memory.
Truck drivers are also not at fault in the large majority of fatal truck-and-car crashes. NTARC's 2025 dataset found heavy vehicles not at fault in 85.7% of fatal heavy-vehicle crashes. Fatigue evidence works both ways. It flags a driver's own risk early, and it protects that driver when the other party was at fault.
What fatigue detection technology does in-cab
In-cab fatigue detection is driver-state AI, not a rear-view mirror with a light on it. The system watches indicators like eye closure duration and head pose in real time. It is tuned to catch the specific signals that precede micro-sleep and drowsy driving, not just obvious signs like nodding off at the wheel.
When those indicators cross a threshold, the system does two things at once. It triggers an in-cab alert to the driver, and it logs the event with a timestamp, GPS position, and speed context for later review. That review is event-based. A manager is not watching hours of continuous footage; they are reviewing a short clip flagged by the AI at the moment the fatigue signal appeared.
This is a meaningfully different model from broader driver monitoring. Fatigue detection specifically targets drowsiness and micro-sleep signals, while wider driver-state AI can also include distraction cues such as looking away from the road, phone-in-hand detection, smoking, and camera obstruction. Both run on the same in-cab hardware but are tuned to detect different risk patterns. See what video telematics means for AU fleets for the fuller category picture.
Vehicle dynamics sit alongside the driver-state signals in the same feed. Harsh acceleration, harsh braking, harsh cornering, and speeding are picked up by the same in-cab system. That matters because fatigue rarely shows up in isolation. A driver approaching a micro-sleep event will often also drift into harsher, less controlled vehicle handling in the minutes before it.
How it ties to HVNL fatigue management evidence
Fatigue detection technology is not a substitute for an operator's fatigue management system. It is a layer that produces defensible evidence for it. Real-time driving-hours and rest-break tracking flags an approaching breach before it happens, rather than after an audit finds it.
That evidence sits alongside, not instead of, an operator's Electronic Work Diary. Crystal integrates with NHVR-approved EWD providers, so work-time and driving-time data flows into the same platform as the fatigue alerts and video evidence. It does not sit in a second, disconnected system the dispatcher has to check separately.
For CoR purposes, this combination, driving-hours evidence plus in-cab fatigue detection plus EWD integration, maps directly to what NHVR audit frameworks expect on fatigue risk. It is evidence of an active safety management system. It is not, and should never be presented as, a legal shield or a substitute for that system.
This matters more as the regulatory window tightens. The Heavy Vehicle National Law Amendment Act 2025 has been assented but is not yet fully in force, with commencement set for 1 August 2026 in the ACT, New South Wales, Queensland, South Australia, Tasmania, and Victoria. Western Australia and the Northern Territory are not covered. In those jurisdictions, the "unfit to drive" duty is set to extend down to all heavy vehicles of 4.5 tonnes and above. Operators with fatigue evidence already flowing into one platform are better placed for that transition than those reconstructing records after the fact.
Crystal Vision as the video layer (exoneration and coaching)
Crystal Vision is the AI video layer that sits alongside fatigue and EWD data inside the Crystal platform. A fatigue alert on its own tells a fleet manager that a threshold was crossed. Crystal Vision's evidence library adds the recorded clip, GPS position, and speed context needed to review what happened.
Every event is auto-classified High, Medium, or Low risk, which lets a safety manager focus coaching time on chronic patterns rather than reviewing every alert manually. Coaching workflows carry an assign, acknowledge, and sign-off trail, so the record of intervention exists alongside the record of the event itself. That same evidence library turns a contested claim into a documented account instead of two conflicting versions of events, with an exoneration pack ready to send straight to an insurer or regulator.
Recording is event-based, not continuous cloud streaming. Local storage keeps a continuous record on the device, but only triggered safety events upload to the cloud, keeping the surveillance footprint to what is needed for review. In NSW and ACT, operators running any camera-based system also carry workplace surveillance notice obligations, including written notice and vehicle signage before the system goes live. This is worth confirming with your compliance team during rollout, not something to market around.
Key takeaways
- Vehicle incidents caused 42% of Australia's 188 worker fatalities in 2024, and human factors, not mechanical failure, drive most serious truck crashes.
- In-cab fatigue detection watches for the physical signs of micro-sleep and drowsiness, and logs every event with a timestamp instead of relying on a driver's memory.
- Fatigue detection is not a substitute for an operator's fatigue management system. It produces defensible evidence for it, alongside EWD data.
- The HVNL Amendment Act 2025 commences 1 August 2026 in six states (not WA or the NT), extending the "unfit to drive" duty to all heavy vehicles of 4.5 tonnes and above.
- Crystal Vision's evidence library and coaching workflows turn a contested claim into a documented account, with an exoneration pack ready for an insurer or regulator.
Frequently asked questions
In-cab AI monitors physical indicators like eye closure and head position throughout the drive. When those indicators cross a threshold that signals drowsiness, the system triggers a real-time alert to the driver and logs the event with video, GPS, and timestamp data for review. It is not a person watching continuous footage. It is an automated detection layer that flags specific moments for review afterward.
The two overlap but are not identical. Fatigue detection is tuned specifically to micro-sleep and drowsiness signals, such as eye closure duration and head pose. Broader driver-state monitoring can also include distraction cues like looking away from the road or phone-in-hand detection. Both run on the same in-cab hardware inside Crystal Vision, but they watch for different risk patterns.
No. Fatigue detection technology provides supporting evidence for an operator's existing fatigue management processes, including EWD records and rest-break compliance. It does not replace the operator's safety management system, and it should never be positioned as a substitute for it or as protection against prosecution. Crystal is the evidence layer. The operator owns the fatigue management system itself.
Book a demo
See how in-cab fatigue detection, EWD integration, and Crystal Vision video evidence work together in one platform.