|Summary / Abstract
||South African underground hard rock mines are typically managed using measurements made daily, weekly or even monthly of key parameters such as face advance, readiness to blast, blast success, temperature or dust levels. Health, safety and productivity can all be greatly improved if a real-time measurement system can inform decision making.
To enable this vision of widespread sensing, communication and decision support, CSIR has developed an open standard architecture for communication of sensor data, and a reference implementation using that standard. The standard, called AziSA, meaning ?to inform? in isiZulu, will allow any device that complies with it to join any network that supports it. The standard is designed for South African gold and platinum mines.
To be practical in a typical Bushveld platinum mine, a real-time sensor system needs to be robust, reliable and cheap:
?Sensors are equipped with power sources that will last as long as the sensor is required. In effect, sensors need to be disposable: cheap enough not to warrant repair if they break.
?Communication in the stope is wireless, to avoid there liability problems associated with wire connections. For example, wires to sensors in the face area are unlikely to survive blasting.
?To get information out of the stope, the power line to the scraper winch is used as a communications link. Unlike a separate communications link, the power line will be repaired if damaged, because of its importance to production.
?Outside the stope, communications are transferred onto the mine communication infrastructure.
To succeed as a management system, the next key element is the use of software decision support tools that will analyse masses of real-time data to extract key parameters and trends for decision makers.
A range of sensors has been developed, including commercial sensors that have had wireless adapters added to make them AziSA compliant, and bespoke sensors that provide low cost sensing of typical mining parameters.
An important parameter in any sensor network is the location of the sensor. This has to be easy and cheap to establish, so a tool has been developed to locate sensors in the confines of a stope, using ultrasonic and radio techniques. Once sensors are installed and located, they associate their location to all data that they produce.
All the data from the system are captured at surface in a flexible database that can be queried by any application that requires data of any kind. It is simple to program alerts, and for actions to be associated with alerts. The use of wireless in stope also allows for communication from surface to the stope, or from stope to stope.
The first implementation of an AziSA compliant system is being installed as part of the Mine Health and Safety Council funded rockfall elimination project. The design is discussed, and a future application for a platinum mine is considered.