How to validate in a variable world: use data, lots of data, both from experiments and from literature

For food safety we want to be absolutely sure! But absolute does not exist; food products, micro-organsims, and humans are biological entities showing large variability. And above that, technical process parameters, also, show variability. In order to validate processes in our variable world, many pieces of data are needed. Data sources include literature data, databases, historical data, experimental data, storage tests, challenge tests, predictive models, basic knowledge, and logic. All these data sources have their weak and their strong points; and therefore, it is useful to draw on of all kinds of sources of data. On one side, due to the large variability in data it is more diffuse and difficult to judge; but on the other side, also more realistic, since in reality, variability does exist.