Introduction Hauling represents 50% percent (Kennedy,1990) of the operating costs in a shovel – truck open pit mining operation, efforts to control or reduce hauling cost in order to reduce their impact over the mining cost has been done along the time. The most common approach is to improve the efficiency of the hauling fleet, reducing idle and waiting times for the trucks by assign them to a loading equipment in real time, this can be done using ad hoc software as [email protected] [email protected] .
These software algorithms work with variables as route length, loading time, hauling time under retain parameters of ore quality and quantity. Based over this information they optimize the variables and assign more or less trucks to a certain face regardless of the truck (as independent unit) performance. Truck Performance We can define the truck performance as the capability of the truck to haul material between two or more points efficiently using all their own mechanical functions and characteristics.
In order to understand this concept we should notice that two trucks, with the same characteristics and age, can have different production rates over the name route; that difference in their production rate should be address to some factors as their maintenance history (which affect their reliability), accidents and driver’s ability. Moreover, these differences could be bigger if we have a fleet composed by trucks from different models and makers. Objectives The objective of this research project is to improve the haul efficiency and reducing cost by develop a tool that can be a complement for the [email protected] [email protected] software .
This tool will select the best trucks, base in their performance, among the let in order to be assigned to the active faces. Methodology Based in the huge amount of information registered by the [email protected] [email protected] software in the mine data warehouse is possible to use this information to find performance patterns, that help us to predict the outcome of some situations and giving us the possibility to analyze and change them. In order to achieve this, the data should be select and process from the data warehouse using data mining tools. The steps in the process to develop this tool are : 1.
Data acquisition : The data will be acquire using data mining from [email protected] or [email protected] database. The data to be collected should describe the code of the truck (model, maker, mine ID) and haul time from any load face to any destination (e. G. Crusher, stockpile,muster). 2 Dataset develop : A new data ease NAS to De develop which associate to each truck a time for a “route”. The number of “routes” will depend of the number of load faces and destinations. Once the “routes” are defined the time that a truck needs to complete a “route” can be calculated from the values in the main database using statistics.
This time value will be assigned to the route” and to the truck. 3. Data processing : Using ANN and based upon production objectives for each load face as output we can determine the best truck combination to be assign to each load face. Evaluation The project objectives has been achieved if we use less trucks to reach the production quotes, which meaner a reduction in the hauling cost and increase of efficiency . In order to improve this tool it has to be capable to be updated also, as a future development we can mix the performance and cost information per truck and use it to assign trucks.