If a mining operation were a human body, then conveyor belts would be the veins.
They carry all the important material to where it is needed in the quickest and most efficient way possible, but what happens when they stop?
A belt breakage instantly stops operations and results in businesses hemorrhaging lost revenues.
Mines typically inspect belt conditions by walking along the belt, but this is a hazardous and time consuming jobs as belt length can be ten kilometers long, and some parts may not be accessible at all.
Miners often outsource belt inspection to specialised companies, with inspection periods varying from a week to even a month.
This approach is expensive and can be unreliable due to the limitations of manual inspection.
Australian company BeltSpy has developed an inspection system that utilises machine vision technology to monitor and maintain conveyor belts in underground mines.
It is designed for use during production hours under full conveyor load conditions.
Beltspy's inspection system is comprised of a high speed digital camera for every belt (with two cameras for double sided installation), a dedicated underground server for each belt, a single workstation, and associated components.
The system uses an image processing algorithm to automatically detect and flag clip joins; splice joins; damaged belt surfaces; and damaged edges, which collectively come under a Points of Interest (POI) umbrella.
The system maintains these precise belt metrics relative to a reference location (RL).
The locations of the POI's distance from the RL are compiled by the system, providing instant access to the POI images.
It also allows for sequential and random belt image browsing.
The single workstation located on the surface allows an operator to inspect the carry and back sides of all belts in the mine by analysing high resolution images anywhere along the belt, and access to images of any flagged locations.
The operator's workstation uses a two-monitor configuration for both carry side belt inspection or a three-monitor configuration if both sides of the belt need to be inspected.
The first monitor is used to control system activity and depicts the graphical representation of the conveyor with all the POI placed on it, while the last two monitors focus on rendering belt fragments (around one metre in length) for carry and back sides, with both images synchronised.
Flagging can be carried out manually or automatically.
An analysis tool allows the potential damage to be flagged during full production periods, allowing for conveyor downtimes to be fully focused on maintenance activities.
It also allows for online monitoring of damage, providing predictive maintenance capabilities.
Operators can also visually compare two images of the same location taken at two different times, so that the user can analyse historical changes in the belt's condition.
The operator can take full belt images at any time and the system can be scheduled to take belt images automatically at pre-defined times every day.
There is no limit of the number of belt condition images that can be stored, with images uploaded in approximately 100 milliseconds.
The BeltSpy system also allows for the monitoring of the development of failing belts in real time until downtimes can be scheduled.