Sustainable value creation by digital predictions of safety performance in the construction industry

The construction industry has, in many years, had a significant higher accident frequency rate than other industries. There is thus an urgent need for knowledge and novel safety management approaches that can contribute to reducing the number of accidents. The project's main aim is to develop knowledge about and methods for application of artificial intelligence in early phases of construction projects to predict future safety performance in the production phase and thus provide improved decision-making support to reduce number of accidents. The project will provide innovation by application of machine learning techniques on available data for novel and proactive safety management approaches. Furthermore, the project will lead to increased sustainability through development of safe and secure working environments. A key foundation in the project is to study project management and safety management as integrated, this also means that not only safety data but data about the project in general will be assessed for application in models and machine learning techniques. The project will be performed in close collaboration with four industrial partners (Sporveien, Skanska, Norconsult and Safetec). The most important R&D challenges addressed are: 1) to demonstrate how success factors in early project phases influence the abilities for safety in production; 2) to explore how machine learning techniques, in combination with risk modelling and computer simulations, can utilize data in projects to provide early warning signals on substandard safety performance in production; 3) to demonstrate how proactive safety management approaches through machine learning can improve hazard control and thereby reduce number of accidents.

The construction industry has, in many years, had a significant higher accident frequency rate than other industries. There is thus an urgent need for knowledge and novel safety management approaches that can contribute to reducing the number of accidents. The project's main aim is to develop knowledge about and methods for application of artificial intelligence in early phases of construction projects to predict future safety performance in the production phase and thus provide improved decision-making support to reduce number of accidents. The project will provide innovation by application of machine learning techniques on available data for novel and proactive safety management approaches. Furthermore, the project will lead to increased sustainability through development of safe and secure working environments. A key foundation in the project is to study project management and safety management as integrated, this also means that not only safety data but data about the project in general will be assessed for application in models and machine learning techniques. The project will be performed in close collaboration with four industrial partners (Sporveien, Skanska, Norconsult and Safetec). The project’s methodology is inspired by action research based on joint processes between researchers and industrial partners to develop unified solutions to specific issues, put the solutions into action and evaluate them. The most important R&D challenges addressed are: 1) to demonstrate how success factors in early project phases influence the abilities for safety in production; 2) to explore how machine learning techniques, in combination with risk modelling and computer simulations, can utilize data in projects to provide early warning signals on substandard safety performance in production; 3) to demonstrate how proactive safety management approaches through machine learning can improve hazard control and thereby reduce number of accidents.