Machine Learning for the prevention of occupational accidents in the construction industry

Purpose and goal

The goal is to strengthen NCC´s development work using ML. In this project, ML has been applied in two areas in NCC: in the field of prevention of occupational accidents and in construction planning using remote sensing. These works are supported and assisted by Professor Shahina Begum, Mälardalen University. NCC has a comprehensive database of registered accidents and incidents that are used as a basis for an ML system. The result shows that the choice of models and validation of data was crucial for the quality i.e. how well the system supports site managers in preventing accidents.

Expected results and effects

The PhD student produced a licentiate report on “Toward Accident Prevention Through Machine Learning Analysis of Accident Reports”. Press release: http://www.byggnorden.se/projekt/mdh-forskning-ska-fa-rebygga-arbetsplatsolyckor-pa-byggen Thesis supervised by Shahina: DEEP LEARNING TO DETECT SNOW AND WATER IN CONSTRUCTION PLANNING USING REMOTE SENSING IMAGES. Publication: Deep Learning in Remote Sensing: An Application to Detect Snow and Water in Construction Sites (Sep 2021) Hamidur R., Mobyen A., Shahina B., Mats F., Adam H. 4th Int. Conference on AI for Industries.

Planned approach and implementation

** Denna text är maskinöversatt ** Work Package (WP) 1: Contribution to method development WP 2: Contribution to a field study by the Health and Safety Domain WP 3: Sparring on data collection WP 4: Assist in validating data quality and accuracy of the cases of use WP 6: Data Analysis Assistance WP 7: Prototype Development WP 8: Test and evaluation

External links

Investigate ML models to prevent workplace accidents in the construction industry.