Towards an autonomous pilot system for a tunnel boring machine

Abstract

Modern Tunnel Boring Machines (TBMs) used in mines or for road construction often consist of several modules for the drill head and further equipment as well as wagons for carrying the mined material (dump). These machines can have a total length of up to 70 meters. In the mines, the machine has to be navigated over long stretches through the mine adits to reach the borehole. The problem is to get there without getting stuck in the galleries. Usually, these long trains are navigated manually and theodolites deal as sensory input for the teleoperator. In this paper we outline a robot system that can deal as a pilot system for such long machines for semi-autonomously driving the TBMs right to their destination. For this, the pilot robot needs to fulfil the following tasks: (1) it has to acquire an accurate 3D map of the tunnel environment, (2) plan a path from a start location to the destination and (3) guide the TBM through the tunnel. We outline the hardware and software system that we envision for our pilot system in this paper. For the hardware, we plan to deploy a Husky A200 platform which is particularly suitable for rugged terrain. For the control software we want to make use of the robot operating system ROS with its many well-implemented modules for map building, navigation and path-planning. As the system is in the planning stage at the moment, we are very interested to learn more about and share experiences and ideas of operating robots in rugged mining environments at the conference.

BIM, machine learning and computer vision techniques in underground construction: Current status and future perspectives

“…Autonomously operating tunnelling systems that aim to remove machine operators from the dangerous underground environment have been realised. One of such solutions is the autonomous pilot system designed for TBM navigation that is capable of on-board map generation and path planning (Ferrein et al, 2012). Other solutions include the BADGER (RoBot for Autonomous unDerGround trenchless opERations, mapping and navigation) that incorporates robotics control techniques, sensor fusion and machine learning (BADGER Consortium, 2017), TBM-cutter changing robot (Yuan et al, 2019), and CPS-enabled autonomous supporting pressure balance control for TBMs (Zhang et al, 2020b).…”