Overview

Smartphones and consumer wearables make it possible to measure health and behaviour day by day in the real world. In recovery settings—such as after knee replacement surgery, activity can change quickly, differ widely between individuals, and fluctuate from one day to the next.

In real-world deployments, wearable and smartphone time-series are often incomplete. Missing days and long gaps can occur for a range of practical reasons, including:

• Non-wear and changes in routine
• Charging and battery constraints
• Bluetooth/app syncing failures
• Software/OS updates and app reinstalls
• Device switching or replacement
• Everyday interruptions (travel, illness, busy periods)
• … and other real-world factors that disrupt collection or transfer

As a result, observed trajectories reflect not only underlying recovery patterns, but also the realities of incomplete observation.

Why missingness from smartphone and wearable matters

Missingness is not always random: gaps can cluster over time, differ across individuals, and relate to behaviour, symptoms, or context. These features can affect summaries, visualisations, and modelling results, so careful methods and clear reporting are essential.

Challenge task

Challenge task details are coming soon.