Document Type
Poster
Start Date
6-3-2026 12:30 PM
Description
Lifelogging — the continuous digital recording of everyday experience through wearable cameras and sensors — has long promised benefits for memory augmentation, health monitoring, and lifestyle analysis. However, current systems follow a capture–store–retrieve model that delays analysis until after data collection, limiting their practical usefulness and creating significant privacy risks through prolonged retention of sensitive raw data. This paper argues that lifelogging should instead analyse data at the moment of capture, enabling timely, context-sensitive support and safer handling of personal information through privacy-by-design. We demonstrate the feasibility of this approach through SelfHealth, an early prototype system built on a custom wearable platform that performs real-time semantic interpretation, on-device anonymisation, and proactive behavioural support. SelfHealth also provides a user-facing mobile dashboard for customised behaviour tracking. Together, these components show that continuous, privacy-aware semantic intelligence can be realised within a practical, end-to-end system.
Included in
SelfHealth: Continuous Semantic Intelligence for Privacy-Aware Lifelogging
Lifelogging — the continuous digital recording of everyday experience through wearable cameras and sensors — has long promised benefits for memory augmentation, health monitoring, and lifestyle analysis. However, current systems follow a capture–store–retrieve model that delays analysis until after data collection, limiting their practical usefulness and creating significant privacy risks through prolonged retention of sensitive raw data. This paper argues that lifelogging should instead analyse data at the moment of capture, enabling timely, context-sensitive support and safer handling of personal information through privacy-by-design. We demonstrate the feasibility of this approach through SelfHealth, an early prototype system built on a custom wearable platform that performs real-time semantic interpretation, on-device anonymisation, and proactive behavioural support. SelfHealth also provides a user-facing mobile dashboard for customised behaviour tracking. Together, these components show that continuous, privacy-aware semantic intelligence can be realised within a practical, end-to-end system.