Generic sleep advice letting shift workers down

By Mary Bolling

Tired Australians are turning to technology for tailored support to improve their sleep habits – but it’s a risky approach for millions of shift workers around the world. 

New CQUniversity research paper says people with unpredictable sleep patterns are getting unhelpful, and potentially unhealthy, advice from one-size-fits-all sleep apps and tech interventions. 

While AI algorithms don’t currently have the data to support shift workers, sleep researchers say interventions designed for shift workers could revolutionise sleep for the at-risk cohort.

The new perspective paper, published in Sleep Advances science journal, says that current consumer AI tools often give shift workers advice that is not just unhelpful, but potentially unsafe.

Lead author and CQU PhD candidate Ruby Smith said smart interventions could level the field for shift workers needing shut eye. 

“Most off‑the‑shelf sleep apps assume everyone goes to bed at night and wakes up in the morning,” she said. “That simply isn’t reality for millions of shift workers, in essential sectors like health care, emergency services, food industries, and freight and logistics.”

“For instance, if a shift worker gets home from night shift and is told to ‘avoid napping’. But shift workers need advice that aligns with their actual lives, not advice designed for 9‑to‑5 schedules.

“Our work shows how artificial intelligence integrated with wearable tech could fill the gap, by directly responding to the unpredictable and variable hours these workers face, and pin-pointing best bedtimes, eating times, exercise windows, and even naps.”

Personalised impact 

The study outlines how tailored AI systems could meaningfully improve sleep by considering factors such as:

• upcoming and previous shift times
• recent sleep duration and quality
• activity levels and heart‑rate data
• fatigue levels reported by the worker
• personal preferences, health conditions and family or social commitments.

Ms Smith said wearable devices aren’t essential for shift workers to benefit from the technology, but they offer real‑time data could make advice even more accurate.

“Large Language Models (LLMs) similar to ChatGPT and Google Gemini can be incorporated into AI sleep interventions and generate ‘intelligent’ responses to human text and speech – essentially meaning its health advice can adapt as users ‘chat’ with the assistant,” she explained.

Flexible fix

While sleep science typically promotes going to bed and waking up at the same time every day, this just isn’t feasible for many shift workers.

But the researchers emphasise that good sleep is still achievable.

Gradually adjusting sleep times before roster changes, managing light and noise during daytime sleep, and using timed physical activity can all help. 

Ms Smith is working with CQUniversity Appleton Institute researchers, led by Associate Professor Grace Vincent, on research to develop a digital sleep assistant for shift workers.

The project is supported by an Australian Research Council (ARC) Discovery Early Career Researcher Award (DECRA).

“This technology has the potential to give the right support, at the right time, in the right context,” Ms Smith said. 

“It could set a new standard for safety, wellbeing and productivity for shift workers.”

CQU’s Appleton Institute delivers a world-leading research program supports the health, well-being and safety of individuals, workplaces and communities.