Machine learning applications to digital rock mechanics data

In rock engineering practice there is often insufficient time, schedule, and computational/expert resources to investigate complex rock mass phenomena in detail. The large amount of multivariate data that characterizes complex rock mass behaviour creates the ideal conditions to use machine learning. Machine learning allows for the use of all relevant data and can inform the design and maintenance of underground excavations such as mines, tunnels, hydropower assets, and utilities.
The purpose of this research is to advance the applications of machine learning in a way that is accessible to practicing rock engineers. To accomplish this, a framework will be developed comprising the acquisition of reliable digital rock mass data and subsequently using it to develop machine learning for forecasting excavation behaviour. The industry partner, RockMass Technologies, is a global leader in the field of digital geotechnical mapping. Currently, RockMass does not offer any machine learning capabilities as part of its product offerings. By participating in this project, RockMass will benefit in-kind from the research advancements made with respect to use of its data in downstream machine learning applications, placing itself at the forefront of machine learning innovation in rock engineering.

Faculty Supervisor:

Davide Elmo

Student:

Partner:

RockMass Technologies Inc.;RockMass Technologies

Discipline:

Engineering

Sector:

Professional, scientific and technical services

University:

The University of British Columbia

Program: