A Path-planning strategy for an Autonomous dozer

In hard rock mine operations, the gravitational segregation of ore particles within a stockpile creates a high variation of particle size. This issue hampers productivity in the comminution process and has a direct effect on the daily output of produced minerals from the mine. Dozer operation can be facilitated to mix the stockpile materials for maintaining a more homogeneous particle size distribution, however this can lead to potentially hazardous situations for the machine and operator due to the dynamic nature of the stockpile. The use of autonomous bulldozers is a new trend for mine equipment operation, improving the safety of dozer operators working in the stockpile by removing them from the machine altogether. This paper describes path-planning strategy for autonomous dozers to operate over a stockpile within a predefined safe region. For this purpose, a stockpile model is developed, where level and particle size properties are used to describe the stockpile in terms of sectors. This information is then used to command the operation of an autonomous dozer over the stockpile based on current SAG mill feed state. Simulation results predict acceptable behavior, provided that sensor information provides enough accuracy to specify safety and performance constrains.