Dynamic blindspots measurement for construction equipment operators. (June 2016)

• Record Type: Journal Article
• Authors: Ray, Soumitry J.; Teizer, Jochen
• Is Part Of: Safety science. Volume 85(2016)
• Page Start: 139
• Page End: 151
• Publication Date: 2016-06
• Subjects: Blind spots from construction equipment and machinery -- Visibility and field of view -- Head posture estimation -- Operator and worker-on-foot -- Random forests and point clouds from range cameras -- Vehicle safety; Industrial accidents -- Periodicals; Accident Prevention -- Periodicals; Safety -- Periodicals; Travail -- Accidents -- Périodiques; 363.11
• DOI: 10.1016/j.ssci.2016.01.011
• Languages: English
• ISSNs: 0925-7535

Abstract

Graphical abstract: Highlights: Limited vehicle operator visibility contributes to many workers-on-foot accidents. Blindspots measurement tests are standardized, but target mostly static blindspots. Body motion and change in operator head posture call for dynamic blindspots tests. Range camera using Random Forests algorithm create real-time dynamic visibility map. Considering operator perspective potentially aids intelligent worker alert systems. Abstract: Limited visibility due to blindspots of construction equipment is responsible for more than half of the visibility-related fatalities in the construction industry. Knowledge of blindspots aids in improving safety on construction sites and the design of the equipment cabin itself. We present a novel technique that allows measuring blindspots dynamically using the head posture of the equipment operator. To achieve this, we perform the head posture estimation with a range camera using Random Forests algorithm. Utilizing the head posture information and the point cloud data of the construction equipment, we generate dynamic blindspots or visibility maps from the operators perspective in real-time. Research findings suggest this approach can potentially be used as an intelligent alert system.