ErgoExplorer: Interactive Ergonomic Risk Assessment from Video Collections

Area of Science:

• Occupational Health and Safety
• Human-Computer Interaction
• Data Visualization

Background:

• Ergonomic risk assessment is increasingly common, yet traditional methods rely on manual observation and limited data analysis.
• Current practices often overlook contextual information and temporal changes, hindering comprehensive risk evaluation.
• Existing methods focus on critical moments, lacking the ability to analyze long-term trends or multiple worker actions.

Purpose of the Study:

• To introduce ErgoExplorer, a novel system for interactive visual analysis of ergonomic risk assessment data.
• To overcome limitations of conventional methods by incorporating contextual information and temporal data from video streams.
• To facilitate the examination of complex relationships in risk assessments across multiple actions and workers over extended periods.

Main Methods:

• Development of ErgoExplorer, a system integrating coordinated multiple views for interactive data analysis.
• Automatic extraction of ergonomic data from video streams.
• Introduction of new visualization views and interactions tailored for domain experts, including integrated score tables.

Main Results:

• ErgoExplorer enables the examination of complex relationships between risk assessments of individual body parts over long sessions.
• The system supports analysis at multiple levels of detail, from overview to individual video frames.
• Demonstrated usefulness of the approach through application to several datasets, highlighting improved analytical capabilities.

Conclusions:

• ErgoExplorer significantly enhances ergonomic risk assessment by providing interactive visual analysis of comprehensive data.
• The system facilitates deeper understanding of ergonomic risks by integrating contextual and temporal information.
• This approach supports domain experts in detailed analysis and exploration of workplace safety data.