Abstract
We aim at an innovative Industial Robot based Automobile Driving Simulator (IRoADSim) to evaluate and forecast intoxicated driver behaviour. IRoADSim includes a system for in-cabin monitoring of drunk driver behaviour (camera, biological sensors), advanced control methods, and a cyber-physical simulator. We will create a comprehensive virtual model of drunk driver behaviours based on synthetic data and real data collected from volunteers.The results of the comprehensive IRoADSim will be evaluated with vehicles, improving the accuracy of the model, thus making the IRoADSim a valid tool for predicting the behaviour of drunk drivers. Ultimately, IRoADSim will support reducing drunk-driving incidents and fatalities.