Taking photos instead of typing

With Farfalle, our approach is that technicians should be able to focus on their core work, while the systems simply work in the background. With Luca’s automated master data capture, service technicians can photograph assets, nameplates, and components. The AI recognizes relevant components, reads technical data, and prepares structured equipment master data for creation in SAP.

The manual capture of new asset components in the field is often time-consuming. Technicians need to identify components, read nameplates, transfer data, and later maintain the information in the SAP system. This creates unnecessary delays and potential sources of error.

This is exactly where Farfalle Luca comes in. From photos of assets, components, and nameplates, the AI recognizes relevant components and extracts technical information. This creates a structured foundation for new equipment master data records in SAP.

The process

The workflow is tailored to everyday work in the field:

1. Take photos
   The technician photographs the asset, individual components, nameplates, and the general condition.
    
2. Start AI analysis
   The app evaluates the images, recognizes relevant objects, and extracts technical information.
    
3. Review results
   The technician remains in control. Recognized objects, assignments, and suggested data can be reviewed, confirmed, or adjusted.
    
4. Prepare SAP data
   The asset is visualized as a logical structure in Farfalle Mobile. The captured information is prepared for transfer to SAP.

Seamless SAP integration and high data quality

The recognized information does not remain isolated. It is integrated into the existing SAP processes.

Farfalle Luca assigns detected components to existing SAP templates. Technical data is mapped to suitable SAP characteristics and can be created as new equipment master records in the backend.

A special focus is placed on data quality, because automated processes should not invent values. Original image information takes priority. Uncertain information is marked or filtered out before it is transferred into the next process step.

Human control

Despite automated AI recognition, the technician remains in control at all times. The recognized results can be reviewed, confirmed, or adjusted before they are adopted. Farfalle Luca clearly shows which components were recognized, which image source the information comes from, and where a component or nameplate was identified in the photo.