Evaluating the effects of safety incentives on worker safety behavior control through image-based activity...

Area of Science:

• Construction Safety
• Behavioral Science
• Computer Vision

Background:

• Construction worker safety is a persistent challenge, with current safety incentive effectiveness difficult to measure accurately.
• Traditional metrics like accident rates don't fully reflect worker compliance with safety regulations.

Purpose of the Study:

• To develop an image-based method for real-time monitoring of construction worker safety behaviors.
• To evaluate the effectiveness of different safety incentive scenarios on worker compliance.

Main Methods:

• Utilized a model integrating OpenPose and spatiotemporal graph convolutional networks to capture worker safety behaviors.
• Assessed safety incentive scenarios involving reward/penalty types and feedback frequency.
• Compared compliance rates for personal protective equipment, hazard avoidance, and safety hook usage.

Main Results:

• Rewarding good compliance is more effective when no feedback on status is provided.
• Penalizing non-compliance is more effective when three instances of feedback are given during tasks.

Conclusions:

• The proposed image-based approach offers a more accurate assessment of safety incentive effectiveness.
• Focusing on observable safe behaviors can promote better safety compliance among construction workers.