Abstract
Reactive and proactive approaches aim to provide dynamic assessment of the overall system risk by following partially different paths. The Bayesian inference-based dynamic risk assessment monitors and processes data from past incidents and near misses. This allows for update of safety barrier failure probabilities in a reactive way. The risk barometer, however, dynamically assesses health of safety barriers on the basis of a wide range of indicators. This allows capturing early deviations and underlying causes of accidents in a more proactive fashion. However, the comparison of the two techniques performed in this chapter shows that they both possess reactive and proactive traits to some extent. Moreover, despite the fact that they may show little complementarity, their overlapping may be useful for (partial) mutual validation.