Automated extraction of hazard information from safety data sheets

There are essentially 3 different processing options for the Safety Data Sheets (SDS) received from suppliers: Print out the data sheet and file it, store it in a digital directory, or read out the contents and store them in a hazardous substances database.

In terms of protecting employees and the environment, the first step should definitely be to read out the SDS digitally. The reason why the data from the SDS has to be checked has been shown by the analysis performed by Clearya: The hazard information in the SDS may be incorrect or incomplete. Even though the supplier of the chemical is responsible for the contents of the SDS, it is also essential for the recipient to check the contents of this data sheet. Only in this way can potential hazards to employees and the environment be ruled out.

But how complicated is the verification of the hazard information in a safety data sheet?

This process follows the data collection from the SDS and involves a semi-automatic procedure. Using existing hazardous substance databases (such as GESTIS or ECHA), classifications and the specialist expert, for example, an initial system can be set up that relates the data and detects possible inconsistencies. This sounds technically extremely complex, but often it is not. In SdbHub, we already do this as part of a downstream checking process for data fields from section 2.