Work

Optimising presence sensor placement in small houses for accurate lighting control with vital sign detection

Year: 2025

Type: article

Abstract: Commercially available presence sensing technologies for smart home lighting are often driven by energy efficiency. However, they struggle to accommodate user’s dynamic behaviours, especially in small houses where multiple activities overlap within compact layouts. Small houses, with mixed-function spaces require sensors that can accurately detect short-range or overlapping activity zones, calling for more user-centric presence sensing that is responsive to user’s behavioural needs and energy efficient. An exploratory study was conducted in a test facility simulating a small house, investigating optimal sensor placement to improve presence detection and minimise unnecessary lighting in unoccupied zones. It examined the performance of 2 conventional and 12 short-range radar sensors, positioned at varying heights and angles. Radar’s potential as a non-wearable sensing solution for health monitoring was also explored. Seventy participants performed three tasks, representing active, sedentary and passive behaviours, under controlled home lighting conditions. Results reinforced that optimal sensor placement in a small house enhanced user presence detection and minimised false triggers. Compared to the conventional sensors, radar-based micro-location sensing presented prospects for accommodating dynamic user behaviours and overlapping activity zones in small houses. It enabled localised lighting based on fine-grain user activity recognition from the respiratory signal, thereby reducing unnecessary lighting energy usage. Notably, it synchronised lighting with user’s ‘sleep-state’ activity and offered a non-wearable solution to monitor user’s health via vital sign detection. These findings lay the foundation for developing adaptive and responsive user-centric home lighting systems for small houses and assistive living environments.

Source: Lighting Research & Technology

DOI: 10.1177/14771535251382820

Authors: R Sithravel, J Landré, A Hurtig-Wennlöf, M Petersson, MBC Aries

Institution: Jönköping University

Language: English

Topic: Context-Aware Activity Recognition Systems

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: hybrid

Estimated APC paid: $4,150