Background
Industrial digitalisation depends on data that organisations are reluctant, or legally unable, to hand over. Operational data reveals production volumes, customer behaviour and vulnerabilities, so the safest choice is often to keep it locked inside one organisation, where it creates far less value than it could.
SID (Säker datadelning för Industriell Digitalisering) worked on the techniques that dissolve that trade-off: differential privacy, federated machine learning and related methods that allow models to be trained and insights to be drawn without the raw data leaving its owner.
