Decision Support Based on Inventory Data

Description

The decisions on the tunnel support in a New Austrian Tunneling Method (conventional drill and blast excavation) for the approaching advance length (round length) are made based upon expert opinions. These opinions depend on the conditions observed on site, tunnel deformations, stresses reached in the lining, etc. All the measurements obtained on-site are recorded and stored for future purposes. Usually, the records created in the early- to mid-digitalization era, are stored as papers or files containing unstructured information (e.g., scanned documents). Such documents may include, e.g., handwritten notes, engineering sketches, printed plots, and photos. Therefore these records cannot be processed easily by a computer or a human.
This paper will discuss a project that aims to find a way to digitalize the valuable data from past engineering projects stored in papers. Even though every tunneling project is different regarding geomechanical boundary conditions, it can be observed that the decisions are often ‘reinvented’ on site, and there were similar project conditions in the past. Obtaining additional records would help automatize the decision-making process, by searching for a past tunneling project under a similar constellation of observed conditions. Upon successfully digitalizing the data, it would present us with a vast amount of data for already completed tunneling projects and the excavation decisions made against its conditions. Analytical approaches aiming to predict rare events might get extra valuable data for models’ validation as extra documents with historical observations will become available in an analysis-ready structural form.
With access to the proposed digitized database of archived projects, it may
also be possible to answer the question of whether the wealth of experience of a
nominated expert can be aided by automatized scripts and learning models.

Period	8 Aug 2023
Event title	22nd Annual Conferencee of the International Association for Mathematical Geosciences: IAMG 2023
Event type	Conference
Location	Trondheim, NorwayShow on map
Degree of Recognition	International