Related Experiment Video
Updated: Aug 5, 2026

Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment
Published on: December 11, 2015
A Feasibility Study of the "Motus" System for Wearable-Based Movement Behaviors at Scale
Nidhi Gupta1, Martin Eghøj2, Tonje Pedersen Ludvigsen1
1National Research Centre for the Working Environment, Copenhagen, Denmark.
Background:
For detailed, large-scale data on 24-hour movement behaviors, we designed a system "Motus" using state-of-the-art wearable and cloud technology, and tested its feasibility on randomly chosen Danish adults in a 2-stage evaluation.
Methods:
Stage 1: We invited 7735 adults, responding to a national occupational health surveillance-2021. Consented participants received a wearable (SENSmotion Plus) and downloaded the Motus app, which provided instructions for wearable attachment on the thigh and for self-reporting work and sleep hours. Following the 7-day measurement, participants completed a feasibility questionnaire. Administrators recorded time spent on Motus-related tasks (eg, postal package preparation). Identified feasibility issues led to revisions of protocol and Motus elements. Stage 2: We invited 6993 adults from a national public health surveillance-2023. Participants used the revised Motus version. We evaluated Motus on the key issues identified from stage 1.
Results:
Stage 1: Feasibility ranged from 77% for social acceptability to 98% for adherence to the measurement protocol. Participants reported spending 73 minutes per week (eg, attaching the sensors) on Motus, while administrators reported 15 minutes per participant. We identified 3 issues: 6% consent rate, 20% lost wearables (but not the data), and 10% wearable patches becoming loose. We addressed these issues by sending reminders, using stronger return envelopes, and replacing patch adhesive with higher quality alternatives, respectively. At stage 2, we observed a higher consent rate (23%) and lower patch complaints (<3%) but higher wearables loss (25%).
Conclusion:
Motus displays promising feasibility for collecting large-scale 24-hour movement behavior data. However, the low participation rate and high sensor loss require improvement before broader implementation, especially in surveillance.

