Student: Dany Labarre
Company: Bentley Systems

Most engineering documents use symbols to characterize the systems they represent. They also include annotations (identifiers, notes, specifications) in textual form to specify certain important properties or identify components. While the association between symbols and components is intuitive for a human expert, this is not the case for a computer.

This project aims to enable the automatic discovery of such association links by a computer, using machine learning techniques. Such an advance could lead to significant efficiency gains and simplified management for companies operating infrastructures that are currently documented manually.