ITEE Innovation Expo 2008 : Project Details

robot path planning using rapidly exploring random trees

Problem Description Background: The multi-boom jumbo rig is a large machine used in underground mining to create access tunnels, or ‘drives’, which are used to obtain access to the ore. Multiple booms are mounted at the front of the machine and are used to drill blast holes into the rock face. These holes are charged with explosives and fired to incrementally advance the tunnel. The shape, size and orientation of the tunnel is determined by the drill hole pattern template being used.

Aim: The aim of this thesis project is to determine which holes of a known blast hole pattern are reachable for the given configuration of the booms. The telescopic booms are capable of blocking each other as the multi-boom jumbo operates in a very small workspace relative to the machine, as such, effective forward planning of the drill sequence is required.

Methodology: The reachable holes are found through the use of a complex path planning algorithm known as Rapidly Exploring Random Trees (RRTs). Planning queries are solved by incrementally building two RRTs of configurations (known to be collision free) at the current hole and destination hole location. The purpose of the algorithm is to establish a connection between the two trees as they explore the space around them.

Conclusions: It has been found that the position of the booms greatly influences the set of reachable holes. By using RRTs, the sequencing of the holes is more effective in that forward planning can determine how the reachable holes for the booms at a given time will be affected as the sequence progresses.