Skip to content

Ein Vergleich

Manual vs. automated data extraction from safety data sheets

  • by
yuriArcursPeopleimages / envato

Capturing data from Safety Data Sheets (SDS) is still a predominantly manual process. We have come up with SdbHub to support business users in this activity.

Manual data extraction from safety data sheets

Copying text from documents is a way that man has probably practiced since shortly after the invention of papyrus. It is also a surprisingly widespread method today. This is because the reason is obvious: the process required for manual typing is very simple to perform.

Another reason for manual entry is that the documents remain within the company’s boundaries and therefore no conclusions can be drawn about the contents of the documents. For certain industries, this secrecy is particularly relevant and also necessary.

Manual data entry is particularly suitable for companies that have a limited number of documents. The condition is, of course, that a pool of employees equipped with the expertise must be available.

Automated data extraction from safety data sheets

Automated data extraction from safety data sheets is particularly relevant for companies that receive several hundred or thousand SDSs annually that need to be captured or updated.

Automated data extraction is independent of staff turnover. Sick waves and an upcoming vacation season has just as little impact on the process.

At SdbHub, we address the concern that SDSs must necessarily, leave the corporate walls with very specific on-premise solutions. Depending on the configuration, all data traces can be deleted after the extraction results have been made available.

A comparison

The advantages briefly described above occur in practice at different companies, which are interpreted differently based on the drive for innovation or digitization.

Error correction:

Error correction in manual capture proves to be very difficult, as the errors made are of a different nature.

With automated capture, the errors tend to be deterministic in nature, meaning they are very similar or likely to always be exactly the same, so adjustments can be made to fix them quickly.

Throughput:

A manual capture of data from an MSDS can take at least several minutes, depending on the requirements. For example, one specialist can process 10-15 MSDSs per hour.

Automatic capture with SdbHub can process 180 SDSs per hour in the standard level and 12,000 SDSs per hour in the high-performance version.

Innovation:

Companies that manually capture several hundred or thousand documents per year tend to have lower levels of innovation, are comparatively less digitized, and perform less well in industry comparisons.

Companies that address the issue of automation are generally better positioned for the future and have a better digitization strategy and are more attractive to skilled workers.

Skilled labor shortage:

Monotonous and highly repetitive activities bring a psychological strain to the work environment that should not be underestimated. Companies that prefer manual processes will be more affected by the shortage of skilled workers and will react particularly sensitively to staff shortages and fluctuations.