A proactive workers' safety risk evaluation framework based on position and posture data fusion

Abstract Construction workers' safety risk evaluation mainly relies on manual observation with safety expert's experience. The manual process is labor intensive and time-consuming. A proactive workers' safety risk evaluation framework is needed to handle this issue. In this paper, the position and posture have been identified as the two key quantitative features, and a position and posture fusion principle is proposed for the evaluation of construction workers' behaviors safety risks. Three participants are invited to involve in the indoor construction activities with working at height three times, and the participants' locations, postures, and video are recorded using sensing devices. The participants' risk levels are evaluated based on posture, location, and posture-location fusion respectively. A comparison of the risk level evaluated based on those three approaches and the ground truth provided by the experts. The results show that the accuracy of automatic safety risk level evaluation based on location and posture fusion is 83%, compared with 81% and 57% achieved by the single-feature based risk level evaluation using location and posture respectively. Thus, the position and posture fusion-based evaluation is more reliable in real situations.

[1]  Faridah Ismail,et al.  Behaviour Based Approach for Quality and Safety Environment Improvement: Malaysian Experience in the Oil and Gas Industry , 2012 .

[2]  Adrian Hilton,et al.  A survey of advances in vision-based human motion capture and analysis , 2006, Comput. Vis. Image Underst..

[3]  Berardo Naticchia,et al.  A monitoring system for real-time interference control on large construction sites , 2013 .

[4]  H. W. Heinrich,et al.  Industrial Accident Prevention: a Scientific Approach , 1951 .

[5]  Heng Li,et al.  Chirp-spread-spectrum-based real time location system for construction safety management: A case study , 2015 .

[6]  Vitor Sousa,et al.  Risk-based management of occupational safety and health in the construction industry – Part 1: Background knowledge , 2014 .

[7]  Nasser Kehtarnavaz,et al.  A survey of depth and inertial sensor fusion for human action recognition , 2015, Multimedia Tools and Applications.

[8]  Miroslaw J. Skibniewski,et al.  Perceiving Interactions on Construction Safety Behaviors: Workers’ Perspective , 2016 .

[9]  Wenbing Zhao,et al.  A Survey of Applications and Human Motion Recognition with Microsoft Kinect , 2015, Int. J. Pattern Recognit. Artif. Intell..

[10]  Heng Li,et al.  Proactive behavior-based safety management for construction safety improvement , 2015 .

[11]  SangUk Han,et al.  A vision-based motion capture and recognition framework for behavior-based safety management , 2013 .

[12]  Maria Pia Cavatorta,et al.  An observational method for Postural Ergonomic Risk Assessment (PERA) , 2017 .

[13]  Lieyun Ding,et al.  Safety barrier warning system for underground construction sites using Internet-of-Things technologies , 2017 .

[14]  V. K. Bansal,et al.  Application of geographic information systems in construction safety planning , 2011 .

[15]  Cristian Sminchisescu,et al.  Human3.6M: Large Scale Datasets and Predictive Methods for 3D Human Sensing in Natural Environments , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[16]  Dongping Fang,et al.  A continuous Behavior-Based Safety strategy for persistent safety improvement in construction industry , 2013 .

[17]  Jimmie Hinze,et al.  Analysis of Construction Worker Fall Accidents , 2003 .

[18]  Jason R. W. Merrick,et al.  Leading Indicators of Safety in Virtual Organizations , 2007 .

[19]  Yantao Yu,et al.  An experimental study of real-time identification of construction workers' unsafe behaviors , 2017 .

[20]  Vitor Sousa,et al.  Risk-based management of occupational safety and health in the construction industry – Part 2: Quantitative model , 2015 .

[21]  Yichen Wei,et al.  Towards 3D Human Pose Estimation in the Wild: A Weakly-Supervised Approach , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[22]  Kathryn Mearns,et al.  Measuring safety climate: identifying the common features☆ , 2000 .

[23]  Vivian W. Y Tam,et al.  Developing a risk assessment model for construction safety , 2010 .

[24]  Qian Liu,et al.  Prospective safety performance evaluation on construction sites. , 2015, Accident; analysis and prevention.

[25]  Jochen Teizer,et al.  Ontology-based semantic modeling of construction safety knowledge: Towards automated safety planning for job hazard analysis (JHA) , 2015 .

[26]  Tao Cheng,et al.  Proximity hazard indicator for workers-on-foot near miss interactions with construction equipment and geo-referenced hazard areas , 2015 .

[27]  Sang Uk Han.  Automated monitoring and systemic analysis of workers' safety behavior in construction operations , 2013 .

[28]  Johnny Wong,et al.  Real-time locating systems applications in construction , 2016 .

[29]  A. Jouybari,et al.  EXPERIMENTAL COMPARISON BETWEEN MAHONEY AND COMPLEMENTARY SENSOR FUSION ALGORITHM FOR ATTITUDE DETERMINATION BY RAW SENSOR DATA OF XSENS IMU ON BUOY , 2017 .

[30]  Lourdes Agapito,et al.  Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[31]  Yuri Odagiri,et al.  Unsupervised Adversarial Learning of 3D Human Pose from 2D Joint Locations , 2018, ArXiv.

[32]  Peter E.D. Love,et al.  A deep hybrid learning model to detect unsafe behavior: Integrating convolution neural networks and long short-term memory , 2018 .

[33]  Seokho Chi,et al.  Relationship between Unsafe Working Conditions and Workers’ Behavior and Impact of Working Conditions on Injury Severity in U.S. Construction Industry , 2013 .

[34]  Heng Li,et al.  Proactive struck-by risk detection with movement patterns and randomness , 2018, Automation in Construction.

[35]  Fernanda Leite,et al.  What is a safe working zone? Designing autonomous monitoring for proactive fall prevention , 2014 .

[36]  Yaser Sheikh,et al.  OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[37]  Helen Lingard,et al.  Behavior-based safety management in Hong Kong's construction industry , 1997 .