Safe Remote Drilling through Predictive Modeling of Hydraulic Hoses

Purpose and goal: ** Denna text är maskinöversatt **

We model flexible hydraulic hoses connected to mining machines to avoid damage during operation. We implemented various physics models and created data-driven models of hose dynamics from synthetic and real data. We generated synthetic data using a new physics model for hose dynamics. Our data-driven model can be used to predict hose movement, track hose condition and control hose shape - making it possible to avoid hose damage during automatic operation.

Expected results and effects: ** Denna text är maskinöversatt **

The scientific result is a data-driven model of hose dynamics that predicts hose movements, the condition of the track hose and controls the hose shape. These results were published in top robotics journals and conferences and a doctoral thesis in 2023. Another results is a physics simulation model of the snake. This model is already used in hose routing design and reduces the need for physical testing. These models can be used to avoid hose damage during automatic drilling operations.

Approach and implementation: ** Denna text är maskinöversatt **

Örebro University (ORU) was responsible for scientific research around data-driven modeling of the hose. Epiroc provided the use case, evaluation metrics and tested the hose routing simulation models. Algoryx supplied the simulation software and developed methods to simulate hoses. ORU and Epiroc were involved in data collection on a real drilling rig. The work was divided into 5 work packages and was carried out over 30 months and coordinated with project meetings.