Mining Goes Digital – Mueller et al. (Eds)
© 2019 Taylor & Francis Group, London, ISBN 978-0-367-33604-2
525
Selection of variables acquired by the on-board monitoring
system to determine operational cycles for haul truck vehicle
Paweł Śliwiński & Marek Andrzejewski
KGHM Polska Miedz SA, Head Quarters M. Skłodowskiej-Curie, Lubin, Poland
Tomasz Kaniewski
OZG Lubin, KGHM Polska Miedz SA, Lubin, Poland
Justyna Hebda-Sobkowicz & Radoslaw Zimroz
Faculty of Geoengineering, Mining and Geology, Wrocław University of Science and Technology,
Wybrzeże Wyspiańskiego, Wrocław, Poland
ABSTRACT: Self-propelled Mining Machines (LHD—Load Haul Dump, Haul Truck) are
the basic technical resource used in copper ore mines in KGHM PM S.A. Machine operation
costs: depreciation, service, consumables, remuneration of operators, etc., account for over
30% of the technical costs of copper ore production. To optimize the cost machines and processes are monitored. The aim of the article is to select variables available in the Haul Truck
monitoring system for operational regimes identification. The development of indicators for
evaluation of the organization of work in the production process, e.g. number of machine
cycles, average cycle time, deviations from the average cycle time, starting/finishing of machine
operation for work shift, etc. may be considered as the first step to optimize the cost of the
process. A method for counting work cycles for LHD loaders has been already developed,
using the pressure signal from the actuator in the hydraulic system. However, in the case of
Haul Truck, the nature of data is different and the new algorithm is needed. Moreover, in
practice, there are limitations in the use of the pressure based indicator due to the loss of data
caused by the pressure sensor’s susceptibility to mechanical damage. Based on these experiences, we will improve the method, thanks to the use of sensors that are not so susceptible to
damage. The article will present the results of analyses regarding the possibility of using other
registered physical variables (engine revolutions, instant fuel consumption, speed, the pressure
in the braking system, etc.), which are much more reliable and resistant to interference.
1 INTRODUCTION
The rapid development of industrial electronics makes monitoring of mining equipment operating underground mines, effective. Mobile machines such as loaders, haul truck, drilling/bolting machines are very specific and their usage and maintenance are demanding. A lot of efforts
have been put in that field [1–6]. The onboard monitoring system is used in underground KGHM
mines in more than 100 mining machines (loaders, truck, drilling/bolting machines). After a
successful pilot phase, a lot of data measured in every 1 s is stored on the machines and transmitted to the data warehouse every day. One might conclude, that data acquisition problem has
been overcome. Unfortunately, the problem of data validation, processing, analysis and finally
reporting are still challenges for such industrial BIG DATA case [7–8]. There are already many
reports prepared on regular basis. However, there is an increasing need for reliable information
for various levels of management staff in the mine. In this paper, we will discuss an important
problem how to automatically transform raw data from haul truck onboard monitoring system
into operational information about production effectiveness, namely number of cycles of haul
truck performed underground in a single shift. This issue has been partially solved for LHD
© 2019 Taylor & Francis Group, London, ISBN 978-0-367-33604-2
525
Selection of variables acquired by the on-board monitoring
system to determine operational cycles for haul truck vehicle
Paweł Śliwiński & Marek Andrzejewski
KGHM Polska Miedz SA, Head Quarters M. Skłodowskiej-Curie, Lubin, Poland
Tomasz Kaniewski
OZG Lubin, KGHM Polska Miedz SA, Lubin, Poland
Justyna Hebda-Sobkowicz & Radoslaw Zimroz
Faculty of Geoengineering, Mining and Geology, Wrocław University of Science and Technology,
Wybrzeże Wyspiańskiego, Wrocław, Poland
ABSTRACT: Self-propelled Mining Machines (LHD—Load Haul Dump, Haul Truck) are
the basic technical resource used in copper ore mines in KGHM PM S.A. Machine operation
costs: depreciation, service, consumables, remuneration of operators, etc., account for over
30% of the technical costs of copper ore production. To optimize the cost machines and processes are monitored. The aim of the article is to select variables available in the Haul Truck
monitoring system for operational regimes identification. The development of indicators for
evaluation of the organization of work in the production process, e.g. number of machine
cycles, average cycle time, deviations from the average cycle time, starting/finishing of machine
operation for work shift, etc. may be considered as the first step to optimize the cost of the
process. A method for counting work cycles for LHD loaders has been already developed,
using the pressure signal from the actuator in the hydraulic system. However, in the case of
Haul Truck, the nature of data is different and the new algorithm is needed. Moreover, in
practice, there are limitations in the use of the pressure based indicator due to the loss of data
caused by the pressure sensor’s susceptibility to mechanical damage. Based on these experiences, we will improve the method, thanks to the use of sensors that are not so susceptible to
damage. The article will present the results of analyses regarding the possibility of using other
registered physical variables (engine revolutions, instant fuel consumption, speed, the pressure
in the braking system, etc.), which are much more reliable and resistant to interference.
1 INTRODUCTION
The rapid development of industrial electronics makes monitoring of mining equipment operating underground mines, effective. Mobile machines such as loaders, haul truck, drilling/bolting machines are very specific and their usage and maintenance are demanding. A lot of efforts
have been put in that field [1–6]. The onboard monitoring system is used in underground KGHM
mines in more than 100 mining machines (loaders, truck, drilling/bolting machines). After a
successful pilot phase, a lot of data measured in every 1 s is stored on the machines and transmitted to the data warehouse every day. One might conclude, that data acquisition problem has
been overcome. Unfortunately, the problem of data validation, processing, analysis and finally
reporting are still challenges for such industrial BIG DATA case [7–8]. There are already many
reports prepared on regular basis. However, there is an increasing need for reliable information
for various levels of management staff in the mine. In this paper, we will discuss an important
problem how to automatically transform raw data from haul truck onboard monitoring system
into operational information about production effectiveness, namely number of cycles of haul
truck performed underground in a single shift. This issue has been partially solved for LHD