Available decision-support tools rarely account for the welfare of technicians in maintenance scheduling for offshore wind farms. This creates uncertainties, especially since current operational limits might make a wind farm accessible but the vibrations from transits might be unacceptable to technicians. We explore technician exposure to vibration in transit based on the levels of discomfort and the likelihood of seasickness occurring on crew transfer vessels (CTVs). Vessel motion monitoring systems deployed on CTVs operating in the North Sea and sea-state data are used in a machine learning (ML) process to model the welfare of technicians based on operational limits applied to modelled proxy variables including composite weighted RMS acceleration (aWRMS) and motion sickness incidence (MSI). The model results revealed poor to moderate performance in predicting the proxies based on selected model evaluation criteria, raising the possibility of more data and relevant variables being needed to improve model performance. Therefore, this research presents a framework for an ML approach towards accounting for the wellbeing of technicians in sailing decisions once the highlighted limitations can be addressed.