Office activity classification using first-reflection ultrasonic echolocation.

Area of Science:

• Human-Computer Interaction
• Biomedical Engineering
• Occupational Health

Background:

• Excessive sedentary time is a significant health risk, particularly for desk-bound professionals.
• Current activity monitoring solutions for office environments are often inadequate.
• There is a need for reliable, contactless technology to track workplace activities.

Purpose of the Study:

• To develop and evaluate a novel echolocation-based system for monitoring activities in an office setting.
• To assess the system's ability to differentiate between sedentary and active tasks at a workstation.
• To provide data for developing interventions to mitigate health risks associated with sedentary work.

Main Methods:

• A contactless sensor array using four first-reflection ultrasonic distance sensors was employed.
• Dispersion-related features were extracted from time-series sensor data.
• Cubic support vector machine classifiers were trained to categorize activities.

Main Results:

• The echolocation system demonstrated the capacity to distinguish between sedentary sitting, typing, writing, and standing.
• Average classification accuracy for sedentary activities surpassed 85%.
• Overall classification accuracy for the tested activity set exceeded 80% in a controlled experiment with six participants.

Conclusions:

• The proposed echolocation-based activity detector is a viable technology for monitoring common workstation activities.
• This non-intrusive method shows promise for promoting healthier work habits and informing targeted interventions.
• Further research can refine the system for broader application in occupational health and ergonomics.