Obtaining a functional product for the retail sector, and in this project the target is the Health and Safety at Work Direction of SONAE MC, to support the management and prevention of accidents and consequent risk reduction (able to identify the probability of future accidents, from historical data).

Develop a local extension of the predictive model to be implemented in 9 units of SONAE MC. The objective is to go beyond reading historical data and create interaction and communication between the predictive model and the local unit, namely provide a pilot product that can be fed through local insights (e.g. failure of a machine) or external (weather forecast).

With the implementation of the developed solution it is intended:

• Reduce by 20% the number of work accidents in the 9 units under study.
• Reduce by 10% the frequency index of work accidents in the 9 units.
• Reduce by 10% the severity index of work accidents in the 9 units.

• A1 – Preliminary studies and technical specifications
• A2 – Design of the prediction model
• A3 – Prototype development in a controlled environment
• A4 – Testing and validation in relevant environment
• A5 – Dissemination, communication and exploitation of results
• A6 – Project management

• An innovative tool supported by ICT tools combined with statistical strategies, data analysis and machine learning algorithms for the early identification of situations with potential risk of accident or development of work-related disease.