Document Template for Preparing the Manuscript

Journal of Mechatronics, Electrical Power, and Vehicular Technology ( 2017 Version )

2 Predictive Maintenance of Heavy Equipment

3 in Mining: Journal Review

4

5 Lufhf Devyna a , Putri Amelya b , Wahyudi Bagus c*

abc
6 Master of Applied Sciences in Manufacturing Technology Engineering,

7 State Polytechnic of Malang

8 Mechanical Engineering Building Jl. Soekarno Hatta No.9, Jatimulyo , Malang City,

9 *Corresponding Author. Tel: +62 817-535-222; Fax: (0341) 404420

10 E-mail: [email protected]

11 Abstract

12 This study give review related the impact of predictive maintenance, especially

13 in industry mining with give maintenance predicitve on the tool heavy. Utilise

14 methodology systematic literature review, This research involve data source from peer-

15 reviewed journals, proceedings conferences and reports industry. This review covers

16 analysis various methods and technologies used, challenges faced, as well as benefits

17 obtained from implementation maintenance predictive with give description practice

18 maintenance in the sector mining. In industry mining, heavy equipment maintenance is

19 key factors for ensure efficiency operations and reduce downtime. This journal analyze

20 various maintenance focused at predictive heavy equipment journal in the mining

21 sector. Predictive maintenance is leveraging approach technology analytics and sensor

22 data for predict failure equipment before it happens, so possible intervention more

23 maintenance appropriate time and effective. Common method used in predictive

24 maintenance including analysis of vibration, thermography, oil analysis, and

25 monitoring real-time conditions using the machine learning and Internet of Things.

26 Apart from that, this journal highlighting challenge implementation predictive

27 maintenance, include will need infrastructure sophisticated technology, availability of

28 quality high data, and competence technical from maintenance staff. With integrate

Page 1of 19 MEV_author_submission_template

 1 findings from various journals, this journal give outlook comprehensive about

2 development current and future predictive maintenance in the mining industry.

3 Keywords: Predictive maintenance; heavy equipment; condition monitoring;

4 machine learning.

5 I. Introduction
6 Predictive maintenance has become focus main in various industries such as

7 aerospace, manufacturing, railways, and energy [1][2] [3] [4], including mining industry

8 because the ability for increase operational efficiency and reduce cost maintenance.

9 Mining industry is one of support development economy national [5]. In mining

10 context, heavy equipment play a role important as tool transport main used for

11 transporting materials from location mining to place processing. The reliability and

12 availability of heavy equipment is very crucial in guard smoothness mining operation.

13 One of very important factor in determine productivity in mining industry is readiness

14 heavy equipment for always operate. Constraint setting happen heavy equipment is not

15 always ready operated and frequently happen damage in a way suddenly or no

16 expected. No damage scheduled this itself very closely connect with operation of units

17 and programs maintenance from that unit [6]. With apply maintenance predictive,

18 mining company can plan maintenance in a way corrective, reducing down time, and

19 increase age use equipment. Predictive maintenance is also becoming more and more

20 get good highlight in practice nor academic because its potential in savings costs,

21 increase safety, and reliability [7]. Development function maintenance possible exists

22 distinction three period in evolution maintenance. The first is corrective maintenance,

23 which is marked with repair only when appear failure. In period This is the approach

24 taken nature responsive, that is action repair done after damage happens, often result

25 time significant downtime and costs high improvement. Second period is preventive

26 maintenance, which is marked with repair planned. In this period, in a way periodic

27 and scheduled based on time or use certain, with objective prevent before the damage is

28 happen [8]. Although more efficient compared to reactive maintenance, this still

29 approach own weakness, that is possibility exists maintenance excessive or replacement

Page 2of 19 MEV_author_submission_template

 1 components are still in condition good. Third period is predictive maintenance, which is

2 marked with inspection, condition monitoring technical, and machine operator

3 participation in maintenance operation [7]. Approach This utilise modern technology

4 like sensors, IoT, and data analysis for monitor condition equipment in real-time and

5 predict failure before happen, can done in a way more appropriate time and efficient,

6 reducing down time and costs maintenance. [9] The goal is for replace traditional

7 maintenance model, however expand it, give more dimensions, smooth, and database

8 based on the framework company maintenance [10].

9 The aim of review journal is discussing applies predictive maintenance on heavy

10 equipment in mining industry. This review covers analysis various methods and

11 technologies used, challenges faced, as well as benefits obtained from implementation

12 predictive maintenance. With understand this study latest in the field is expected can

13 give more good insight about development and implementation of maintenance

14 strategies more predictive effective in mining industry.

15 II. Materials and Methods

16 The most important concepts developed in predictive maintenances are Reliability

17 Centered Maintenance and Total Productivity Maintenance [11]. Reliability centered

18 maintenance progams in the mining industry enabling maintenance engineers to choose

19 the best mix of plans and policies for maintenance based on cost considerations is the

20 aim of RCM in the mining industry. This comprises two stages; critical analysis and

21 failure mode identification, together with maintenance chores and options. The

22 possibility that the expense will occur or the extent of the damage are typically the

23 bases for severity.

24 The models developed for equipment maintenance function as diagnostic tools as

25 well as prognosticators. According to the body of existing research, maintenances

26 model can roughly divided in 3 categories:

Page 3 of 19 MEV_author_submission_template
 1 a) Data drivens model: Large datas set are used in data-driven maintenance models.

2 The next step is to make predictions by differentiating patterns and relationships.

3 This models adaptability and training on large data sets are what make it unique.

4 After that, fresh data based on discovered patterns [12]. A number research

5 conducted by Alsina et al (2017), Ali et al (2016), and Y. Wang et al (2024) adopt

6 approach based on this data for engine prognosis and diagnosis. In contrast to

7 physics based on models, this model show superior accuracy in identification

8 point degradation, thereby expand the models general application [13] [14] [15].

9 b) Physics based models: This models use formal mathematics and fundamental

10 physics to forecast ascertain outcomes. This model is highly accurate, but if the

11 accurate, but if the failure mevhanism is only partially known, its usefulness will

12 decrease.

13 c) Knowledge based models: Model based is bsuilt on previously acquired

14 knowledge. Expert knowledge is available to help with forecasting and decision

15 making. This models then forecast or decide on the basis of new information, take

16 into account the skill domain that is particular to a rule or decision [16]. Most of

17 the models are in category this, like "system experts” and “fuzzy logic”, imitating

18 cognitive processes man exemplify application of the model in analysis signal

19 from bearing rollers and wheels tooth [17]. Although this models proven effective,

20 ambiguity at expert knowledge and the real world can hinder taking decisions.

21 Models that rely on knowledge and physics need in-depth details about the machine

22 and its surroundings. Models used for prognosis or diagnosis benefit from this

23 accumulated information [18]. The complexities of knowledge-based and physics are

24 fully explained in [19]. The three afrementioned paradigmas and the underlying sub-

25 models that underlie them are depicted in Figure 1, outlining the model selection

26 criteria [20]. Promoted the fusion of data-drivem and physics-based approaches to

27 increanse the preditictive model’s capcity to generalize when deisgned for industrial

28 machinery. Similarly, for fault diagnosis proposes merging knowledge-based and data-

29 driven approaches [21]. As shown in Figure 1, while data-driven paradigms depen on

30 Machine Learning and Deep Learning, other method include statistical and stochastic

Page 4of 19 MEV_author_submission_template

 1 algorithma (e.g., trends extrapolations, PCA, Multi refime, ARMA, Regresion), as well

2 as Bayesian networks, particle filters, and kalman filters [22].

4 Figure 1 Details various approach applicable modeling for Predictive Maintenance

5 2.1. Maintenance Models in Mining

6 Reability centered maintenance and condition monitoring based have historically

7 been the most popular maintenance-related progams in the mining industry [23].

8 2.1.1. Condition Based Monitoring or Condition Monitoring

9 While there are minor variations, studies like Benbouzid (2021) and Aydemir (2019)

10 have categorized and described predictive maintenance as condition monitoring in the

11 same way. Processes that reliably record and track the present state of operation of a

12 specific machine component, subsystem, or system are a part of condition monitoring

13 [24] [25]. Real-time situations involve these data streams being connected to alarms

14 there inform maintenances teams deviations for standard operating procedures or

15 manufacturer standards. Their data is most thoroughly useful for diagnosing

16 equipment [26] [27] [28].

17 Predictive maintenance goes a step further and inccorporates a predictive element

18 into the caracteristics of condition monitoring maintenance. Condition based

19 monitoring has proven successful in certain application domains, but it methods for

20 data processing and analysis are constantly in need of enhacement [29]. Condition

21 monitoring data is typically a helpful source when creating a system or model for

Page 5 of 19 MEV_author_submission_template
 1 predictive maintenance. Two types of software are utilized in mining; Condition

2 monitoring system and computer maintenance management systems. Information

3 including manufacture recommendations, maintenance activity history, and associated

4 events are store in computer maintenance management systems. However, critical

5 prhysical parameters like pressure, temperature, vibration, acoustics, etc. are stored in

6 condition monitoring system [30].

7 Conventional condition monitoring systems employ one, a few, or all of that

8 following strategies. These include of temperature monitoring, torque measurement,

9 debris oil analysis, vibration monitoring, electrical analysis signal, acoustic emissions

10 [31], and SCADA signal [32] List a few of the concerns or difficulties that were

11 identified from the information gathered during the afotementioned procedure [33]

12 demonstrates that the benefit of advanced condition monitoring is similar to that of

13 anomaly detection. It is only a matter of detecting the related occurences and notifying

14 the operator since the second purpose is to identify an event or sequence of event that

15 deviates from usual working conditions.

16 2.1.2 Reability Centered Maintenance

17 When the Boeing 747 and its more shophisticated electronics were first constructed

18 and made available for purchase in 1960, reliability-centered maintenance (RCM) was

19 first developed and implemented [34]. Nevertheless, the mining industry did not

20 embrace RCM until 1988, a delay of more than 20 years [35]. This method uses a

21 statistical reliability model that is based on probability to identify the characteristic of

22 equipment downtime. In doing reliability analyses [36], emphasized the significance of

23 probability dostribution and statiscal ideas.

24 Research has been done on this strategy. RCM is used in mining to assist

25 maintenance engineers in selecting the best combination of plans and policies for

26 maintenance, taking cost considerations into account. Usually, there are two phases

27 involved; identifying the maintenance alternatives and duties, and conducting a failure

28 mode and critical analysis. The likelihood of an event occurring and the extent or

29 expense of the harm are typically the foundations of a criticality analysis [38].

Page 6of 19 MEV_author_submission_template

 1 The benchmarks for assessing the efficiency of any equipment maintenances progam

2 or activities consist of two essential elements. Specifically MTBF (Mean Time Between

3 Failure) and Overall Equipment Effectiveness (OEE). Whereas the MTBF indicator

4 gauges avaibility, the OEE indication gauges efficiency. In an attempt to characterize the

5 dependability of surface mining machinery [38], listed four potential data processing

6 methods for efficien RCM. To get the best RCM outcomes, it also covers when and how

7 to utilize it. Furthurmore, this study determined that the following information should

8 be gathered in order to complete the suggested data analysis procedures:

9 2.1.3 Statistic Methode for Modeling Equipment Data Failure

10 As the RCM section discusses, the cost function is crucial in deciding on maintenance

11 plans and policies. Consequently, the link between characteristics like maintenance

12 costs, the cost of lost production during downtime, and total production costs is

13 typically displayed by standard statistical models and distributions. Furthermore, the

14 life cycle of an item of equipment can be ascertained by statistically analysing

15 maintenance data. While statistical methods are helpful, choosing a statistical

16 distribution model often involves making unwarranted assumptions, which is one of its

17 disadvantages. For instance, it is expected that after each repair, when using a statistical

18 stationary model distribution, a machine or any of its parts will be in the same state as

19 when it was originally manufactured.

20 To simulate data on equipment failure, statistical functions such as the Weibull,

21 exponential, and lognormal distributions are commonly employed. They offer

22 conventional methods for figuring out the distribution of function or reliability. Weibull

23 statistical distribution helps pinpoint malfunctions and enhance the dependability of

24 dewatering pumps and compressors in mines [37]. Further demonstrated that machine

25 learning approaches can be used to produce more accurate reliability. These studies

26 also emphasise the subsequent stages that organise the flow of the Predictive

27 Maintenance system design process.

28 2.3 Subsystems and Machines

29 A machine frequently has several subsystems, subassemblies, and components. This

30 means that this sub-hierarchy level directly and directly affects the state or condition of

Page 7 of 19 MEV_author_submission_template
 1 the entire machine. Therefore, any necessary maintenance must give priority to the

2 maintenance needs of sub-level components because the machine's overall working

3 status is established by aggregating the status of all of its parts and components. To

4 make this happen, though, each tier of the hierarchy needs to contain the following:

5 digital data about the failure modes of components, operational objectives of individual

6 components or groups of components, and information about component identification

7 and failure modes. Next, utilising fault frequency and fault duration (time) as

8 classifiers, Figure 2 provides a graphic representation of the classification of equipment

9 parts, maintenance procedures, and maintenance decision making. Predictive

10 Maintenance works well with components that are more critical and have longer

11 downtime and fewer failures per unit. Regular types of maintenance should be

12 performed on components with low downtime and low frequency of errors or failures;

13 Predictive maintenance should be applied to components with higher frequency of

14 errors or failures and lower downtime.

15

16 Figure 2 Classification component based on frequency component interruptions and downtime

17 III. Results and Discussion

18 3.1 ARM with similarity semantics

Page 8of 19 MEV_author_submission_template

 1 According to literature, integration similarity semantics is very important when face

2 to face with text data, and various method used for measure similarity semantics

3 between the words or text. One of the method is with use word embedding, which

4 changes a word becomes vector that can compared to in other words and this is a very

5 interesting field mining .

6 3.2 Predictive maintenance techniques in management oil lubricant

7 Predictive maintenance of lubricants with sample collected oil from heavy

8 equipment a classified to in three group that is can reported, critical and routine. This

9 study show that potency leaks, hydraulic filters, and other abnormal events must

10 monitored and repaired in condition closed. His height tie the contaminant being

11 measured caused by failure function moment charging oil or testing oil. Therefore that's

12 it in matter This microfitration oil is highly recommended for reduce stay impurity and

13 prolong age equipment.

14 Integrating machine learning predictive maintenance methods in mining can drive a

15 revival and change the contours of operational efficiency, safety and sustainability in

16 this industry. Machine learning algorithms can predict equipment failures with high

17 accuracy, minimize unplanned downtime, and optimize maintenance schedules. Some

18 methods use prognostic and health management but do not concentrate on tracking

19 engine vibration and how it affects the driver. Moreover, current solutions are offline

20 based, unable to scale, and unable to operate in real-time. The residual usable life (of

21 dumpers can be estimated by assessing engine and seat vibration using a cloud-based

22 condition monitoring system.

23 3.3 Machine Learning at Predictive Maintenance

24 Machine Learning can applied to fresh deep learning expositions at different steps,

25 including as evaluation, modelling, and data-related (data preparation and

26 interpretation) stages. Machine learning algorithms can be employed to build a

27 predictive model that predicts whether a certain piece of equipment will succeed or fail.

28 Machine Learning algorithms can be employed in the evaluation step to evaluate these

29 models' performance and pinpoint areas in need of development. All things considered,

Page 9 of 19 MEV_author_submission_template
 1 the Fresh deep learning architecture's Predictive maintenance systems accuracy and

2 performance can be enhanced with the application of machine learning.

3
4

5 Figure 3 Flowchart showing what is the Predictive Maintenance model can chosen in a way effective

6 in relationship with Machine Learning methods

7 In today’s situation, information is often checked incorrectly, not complectely or only

8 partially and errors rarely occur. To overcome these difficulties, scientists, for example,

9 [39] and [40] have planned machine learning based models [41]. Developed a unique

10 stage using unaided understanding, using a semi – regulatory learning approach for the

11 caracterization of irregularities for collections of information that are to some extent

12 called. They obeserved that this approach is promising because it can improve

13 multiclass ordering, while preventing overfitting that occurs due to imbalanced class

14 information.

15 3.4 Broad Types of Machine Learning

Page 10of 19 MEV_author_submission_template

 1 On the lower floor two classifications of machine learning the machine learning

2 tradition approach and the deep learning approach are covered in great detail in this

3 study. The goal of this study is to point several applications for it.

4 3.4.1 Traditional machine learning-based techniques

5 Shallow neural networks are another name for traditional machine learning. The most

6 popular traditional machine learning techniques in predictive maintenance mining are

7 Decision Trees, K-Closset Neighbour (KNN), Backing Vector Machine (SVM), Irregular

8 Woodeep learningand (RF), and Fake Brain Organisation (ANN). Furthermore, List of

9 machine learning approaches includes machine learning algorithms that are frequently

10 employed in the mining sector for fault analysis, RUL prediction, and operational and

11 failure mode detection.

12 3.4.2 Deep learning-Based Procedures

13 Deep learning-based Procedures or Profound Progessing or called deep brain

14 organization (DNN), have common executions in the endeavor, for example component

15 and patterm learning, extraction, ID and weakness mode setting, and Rul expectation.

16 There are a number of methods that can be used, deep belief, inclyding autoencoding,

17 recurrent neural, convolutional neural, and reinforcement learning. Deep learning

18 computing is invaluable for handeep learning complex, multi-layered information

19 collection and can achieve leading-edge execution in a variety of applications.

20 3.4.3 Hybrid Model

21 This segment investigates the use of machine learnings in Predictive Maintenance for

22 mining businesses and its true capacity in further developing the dependability of

23 equipment and supporting system. Big despite the fact that predictive maintenance is

24 based on machine learning, there are still a number of challenges, including incomplete

25 information, unteliable information, and unsatisfactory information. Machine learning’s

26 goal is to consolidate the family and the intrinsic conditions so that thet can be shared

27 with mining. Optimising techniques and models for anomaly and defect detection in

28 diverse mining stages and segment-specific contexts is crucial to realising the full

29 potential of machine learning-based predictive maintenance in the mining industry. As

30 a result, quality and operational efficiency will be highlighted.

Page 11 of 19 MEV_author_submission_template
1

2 Figure 4 Predictive Maintenance Process traditional MACHINE LEARNING based

4 Figure 5 Predictive Maintenance Process classic DEEP LEARNING based

6 Figure 6 Modification Structure 3 Phase Predictive Maintenance [13]

7 This summation uses an overall approach to evaluate the perfomance of the

8 predictive maintenance model in terms of loaction and determination. They examine

Page 12of 19 MEV_author_submission_template

 1 the utilization of machine learning computing for discovery and determination of

2 problems in mining equipment, establishing a three-stage cycle for success.

3 3.5 Future Researches Direction

4 Reception & adequacy prescient upkeep models, especially at areas like the mine

5 business, relies upon their capacity to be vigorous and versatile. Key areas of future

6 examination that are significant for futher developing prescient support models. These

7 regions incorporate the monetary posibility of that models, their capacity to sum up

8 across various unit inside an industry, potential from expanded prescient abilities using

9 numerous information types, and the fuse of continuous dynamic cycles.

10 3.5.1 Improving Predictive Maintenance Models

11 Predictive maintenance models/ economic feasibility is an important factor to take

12 into account. A thorough cost-benefit analysis can validate the model’s viability and

13 cost-effectiveness by taking into consideration possible cost savings, ongoing

14 maintenance expenses, and initial setup costs. A solid financial understanding of the

15 models application in the mining industry may be developed by taking into account

16 factors like return on investmen, hardware cost, maintenance requirements, and the

17 influence on operational efficiency and reduced downtime. The capacity of predictive

18 maintenance models to generalize and adapt to various circumstances is a crucial

19 component in their improvement. Although the study’s current focus is on particular

20 mining sector components, predictive maintenance models can also be advantegous to

21 other industry groups. Future studies ought to evaluate how generalizable these.

22 3.5.2 Improve Predictive Ability

23 Predictive maintenance models, like the ones examined in this studyt, are frequently

24 developed using structured data, like data sensor and maintenance log. Unstructure

25 datas, on the other hand, can also offer insightful information. Examples of this type of

26 data include text from maintenance reports, and sensor logs. By incorporating natural

27 languages processing methods for manage this unstructures input, the model’s

28 predictive power can be greatly enhanced. Future studies should take into other

29 variables that could affect the rate of equipment failure, such as the state of weather,

30 consumer demand, or supplier perfomance.

Page 13 of 19 MEV_author_submission_template
 1 The capacity of a predictive maintenance model to manage the uncertainty present in

2 real-world situations is another factor that determines how effective the model is.

3 Another area that has to be addressed is long-term perfomance evaluation. In order to

4 forecast equipment failures and plan maintenance chores, predictive models frequently

5 rely on past data. Variations in operating circumstances, age of the equipment, and

6 maintenance interventions can all affect how accurate these projections are. Model

7 acuracy can be increased by conducting through long-term perfomance assessments,

8 which entail tracking real equipment perfomance over time and constrasting it with

9 model projections. A careful examination of the body at researches at predictive

10 maintenance on the mine sector has shown that the field has made considerable strides

11 and continues to do so. This paper offer a thorough synopsis of the state of predictive

12 maintenance, emphasizing its applications, ongoing investigations, and prospects for

13 the future. This heterogeneity may not be well capture by deterministic assumptions

14 about failure likehood. Enchansing the model with risk management and uncertainty

15 analysis methods will result in a stronger foundation for making decisions. This

16 strategy has the potential to enchace predictive skills even more when combined with

17 an enchanced feature engineering procedure.

18 IV. Conclusions
19 The more lots of data available for management facilities, however sector not yet

20 mandate potency full of what it offers. For example, predictive maintenance is a rare

21 area considered by the company manager facilities, one possibility the reason is Because

22 lack of structured data and explicit labels. By standardizing oil collection intervals,

23 maintenance managers can use trend lines to predict equipment behavior over the life

24 of the equipment. To increase equipment reliability, this technology should be used

25 more widely in land logistics, land construction and heavy vehicles in general.

26 This paper offers a thorough sypnosis of the state of predictive maintenance,

27 emphasizing its applications, ongoing investigations, and prospects for the future.

28 Predictive maintenance has gained traction in the mining sector because it may

Page 14of 19 MEV_author_submission_template

 1 minimize downtime and improve maintenance planning, which boosts productivity

2 and competitiveness.

3 Artificial intelligence approaches, particularly machine learning algorithms, have

4 become vital tools in predictive maintenance. These models have demonstrate efficiency

5 in identifying precursors to equipment malfunction and forecasting equipment failure,

6 thus enhancing operational effectiveness. This has allowed for the examination of the

7 efficacy of various machine learning methods models such as neural networks, decision

8 trees, linear models, and their hybrids have proven to be highly effective in forecasting

9 machine failure and remaining useful life.

10 Predictive maintenance model implementation is not whitout challenges, despite its

11 apparent advantages. Effective data administration procedures are essential as it has

12 been highlighted how crucial feature engineering, preprocessing, and high-quality data

13 are to precise forecasts. Further study has been identified in the areas of implementation

14 cost-effectiveness, external influences, unstructured data inclusion, and the

15 requimenent for models to account for uncertainty in the actual world.

16 The literature demonstrates different ways according to the type of equipment, the

17 type of failure, and the availability of data. Nonetheless, data indicates that predictive

18 maintenance is not only economical but also practical in the mining sector, which

19 supports more study and advancement in this field. When assessing predictive

20 maintenance as it is now employed in the mining sector, it is evident that although

21 great strides have been made, agreement on the best practices and models has not been

22 established. The literature demonstrates different ways according to the type of

23 equipment, the type of failure, and the avaibility of data. Nonetheless, data indicates

24 that predictive maintenance is not only economical but also practical in the mining

25 sector, which supports more study and advancement in this field.

26 References
27 [1] O. Dayo-olupona, B. Genc, T. Celik, and S. Bada, “Adoptable approaches to

28 predictive maintenance in mining industry : An overview,” Resour. Policy , vol. 86,

29 no. PA, p. 104291, 2023, doi: 10.1016/j.resourpol.2023.104291.

Page 15 of 19 MEV_author_submission_template
 1 [2] T. Berghout, L. Mouss, O. Kadri, L. Saïdi, and M. Benbouzid, “Engineering

2 Applications of Artificial Intelligence Aircraft engines Remaining Useful Life

3 prediction with an adaptive denoising online sequential Extreme Learning

4 Machine,” Eng. Appl. Artif. Intel. , vol. 96, no. September, p. 103936, 2020, doi:

5 10.1016/j.engappai.2020.103936.

6 [3] G. Nota, A. Postiglione, and R. Carvello, “Text mining techniques for the

7 management of predictive maintenance,” Procedia Comput. Sci. , vol. 200, pp. 778–

8 792, 2022, doi: 10.1016/j.procs.2022.01.276.

9 [4] ZM Çınar, AA Nuhu, Q. Zeeshan, and O. Korhan, “Machine Learning in

10 Predictive Maintenance towards Sustainable Smart Manufacturing in Industry 4 .

11 0,” no. Machine learning, 2020.

12 [5] DN Daman Agus, "ANALYSIS OF OVERALL EQUIPMENT EFFECTIVENESS

13 (OEE) ON THE HITACHI EX2500-6 EXCAVATOR," vol. 1, no. 6, pp. 847–855,

14 2020, doi: 10.31933/DIJEMSS.

15 [6] F. Widjaja, D. Hendriana, and H. Nasution, "Predictive Maintenance of Heavy

16 Equipment in Indonesia Leading Heavy Equipment Company," pp. 24–30.

17 [7] T. Dayo-olupona, “Adoptable Approaches to Predictive Maintenance in Mining

18 Industry : An Overview,” Resour. Policy , vol. 86, no. PA, p. 104291, 2023, doi:

19 10.1016/j.resourpol.2023.104291.

20 [8] P. Odeyar, DB Apel, R. Hall, B. Zon, and K. Skrzypkowski, “A Review of

21 Reliability and Fault Analysis Methods for Heavy Equipment and Their

22 Components Used in Mining,” pp. 1–27, 2022.

23 [9] S. Škerlič and E. Sokolovskij, “Analysis of Heavy Truck Maintenance Issues,” vol.

24 34, pp. 24–31, 2020.

25 [10] O. Alshorman et al. , “A Review of Artificial Intelligence Methods for Condition

26 Monitoring and Fault Diagnosis of Rolling Element Bearings for Induction

27 Motors,” vol. 2020, no. CM, 2020.

28 [11] A. Vorkapić, P. Kralj, and D. Martinović, “The Analysis of The Maintenance

29 Systems of an LPG Carrier's Liquefaction System Main Components,” vol. 31, pp.

30 3–9, 2017.

Page 16of 19 MEV_author_submission_template

 1 [12] O. Bektas, J. Marshall, and JA Jones, “Comparison of Computational Prognostic

2 Methods for Complex Systems Under Dynamic Regimes : A Review of

3 Perspectives,” Arch. Comput. Methods Eng. , vol. 27, no. 4, pp. 999–1011, 2020, doi:

4 10.1007/s11831-019-09339-7.

5 [13] EF Alsina, M. Chica, and K. Trawi, “On the use of machine learning methods to

6 predict component reliability from data-driven industrial case studies,” 2017, doi:

7 10.1007/s00170-017-1039-x.

8 [14] D. Ali and S. Frimpong, “International Journal of Industrial Ergonomics Meaning

9 of intelligence models for predicting the performance of hydro-pneumatic

10 suspension struts in large capacity dump trucks,” Int. J.Ind. Ergon. , vol. 67, no.

11 June, pp. 283–295, 2018, doi: 10.1016/j.ergon.2018.06.005.

12 [15] Y. Yu, Y. Wang, Q. Li, and B. Jiao, “A Predictive Energy Management Strategies

13 for Mining Dump Trucks,” 2024, doi: 10.32604/ee.2023.044042.

14 [16] E. Apriaskar, D. Prastiyanto, A. Baskoro, V. Lokahita, and C. Photong, “Journal of

15 Mechatronics, Electrical Power, Genetic algorithm-enhanced linear quadratic

16 control for balancing bicopter system with non-zero set point, ” vol. 14, no. 2, pp.

17 105–113, 2023.

18 [17] Arpan Chaudhary, "Developing Predictive Models for Fuel Consumption and

19 Maintenance Cost Using Equipment Fleet Data," 2019.

20 [18] A. Kahraman, “ThinkIR : The University of Louisville's Institutional Repository,”

21 2018.

22 [19] BP Iskandar, A. Cakravastia, US Pasaribu, and H. Husniah, "Performance-Based

23 Maintenance Contract for A Fleet of Dump Trucks Used in Mining Industry," pp.

24 117–122, 2014.

25 [20] TP Carvalho, FAAMN Soares, R. Vita, and P. Francisco, “Computers & Industrial

26 Engineering A systematic literature review of machine learning methods applied

27 to predictive maintenance,” Comput. Ind. Eng. , vol. 137, no. April, p. 106024, 2019,

28 doi: 10.1016/j.cie.2019.106024.

29 [21] Q. Zhou and P. Yan, “Research on a configurable method for fault diagnosis

30 knowledge of machine tools and its application,” no. 174, pp. 937–960, 2018.

Page 17 of 19 MEV_author_submission_template
 1 [22] R. Keith Mobley, “Predictive Maintenance. In: Plant Engineer's Handbook,” pp.

2 867–888, 2001, doi: 10.1016/B978-0-7506-7328-0.50052-5.

3 [23] MM Sytsems and MW Lewis, “Maintenance of mobile mining equipment in the

4 information age,” vol. 7, no. 4, pp. 264–274, 2001.

5 [24] M. Benbouzid, T. Berghout, N. Sarma, and S. Djurovi, “Intelligent Condition

6 Monitoring of Wind Power Systems : State of the Art Review,” 2021.

7 [25] G. Aydemir and K. Paynabar, “Image-based Prognostics Using Deep Learning,”

8 IEEE Trans. Ind. Informatics , vol. PP, no. XX, p. 1, 2019, doi:

9 10.1109/TII.2019.2956220.

10 [26] R. Bodo, M. Bertocco, and A. Bianchi, “Smart Maintenance Applications,” pp. 27–

11 32, 2020.

12 [27] C. Model, H. Wiemer, and L. Drowatzky, “applied sciences Data Mining

13 Methodology for Engineering Applications ( DMME )— A Holistic Extension to

14 the,” 2019.

15 [28] D. Ali and S. Frimpong, "Artificial intelligence, machine learning and process

16 automation: existing knowledge frontier and way forward for mining sector,"

17 Artif. Intel. Rev. , no. 0123456789, 2020, doi: 10.1007/s10462-020-09841-6.

18 [29] T. Zhu, Y. Ran, X. Zhou, and Y. Wen, “A Survey of Predictive Maintenance :

19 Systems, Purposes and Approaches,” no. Deep learning, pp. 1–38.

20 [30] AS Adigama and GD Haryadi, "Reliability analysis and critical part lifetime

21 prediction of heavy duty articulated dump truck on coal mining industry.

22 Reliability Analysis and Critical Part Lifetime Prediction of Heavy Duty

23 Articulated Dump Truck on Coal Mining Industry," vol. 030088, no. April, 2019.

24 [31] FL Bronté, CR Pagotto, V. Bicalho, and C. De Almeida, “Predictive Maintenance

25 Techniques Applied in Fleet Oil Management Predictive Maintenance Techniques

26 Applied in Fleet Oil Management,” no. October, 2016.

27 [32] N. Gomolla, J. Hebda-sobkowicz, A. Michalak, and J. Wodecki, “Predictive

28 Maintenance of Mining Machines Using Advanced Data Analysis System Based

29 on the Cloud Technology Predictive Maintenance of Mining Machines Using

30 Advanced Data Analysis System Based on the Cloud Technology,” no. January

Page 18of 19 MEV_author_submission_template

 1 2020, 2019, doi: 10.1007/978-3-319-99220-4.

2 [33] B. Madhu et al. , “Predictive Maintenance of Mining Haul Truck Engines Using

3 Oil Sampling and Telemetry Data,” vol. 10, no. 9, pp. 56–63, 2020.

4 [34] C. Yuan, “Predictive Analysis on Maintenance of Mining Dump Truck,” vol. 340,

5 pp. 848–851, 2013, doi: 10.4028/www.scientific.net/AMM.340.848.

6 [35] E. Siami-irdemoosa and SR Dindarloo, “Prediction of fuel consumption of mining

7 dump trucks : A neural networks approach,” vol. 151, pp. 77–84, 2015.

8 [36] E. Angeles and M. Kumral, “Optimal Inspection and Preventive Maintenance

9 Scheduling of Mining Equipment,” J. Fail. Anal. Prev. , 2020, doi: 10.1007/s11668-

10 020-00949-z.

11 [37] A. Afiyuddin and A. Sudiarno, "Through Contract Models or Company it Self".

12 [38] S.R. Simard, M. Gamache, and P. Doyon-poulin, “Development and Usability

13 Evaluation of VulcanH , a CMMS Prototype for Preventive and Predictive

14 Maintenance of Mobile Mining Equipment,” pp. 326–351, 2024.

15 [39] NB Laulita, "Journal of Industrial Systems Optimization How Organizational

16 Culture Moderates the Effect of Total Productive Maintenance Practice on

17 Operational Performance? Evidence from Indonesian Mining Industry,” vol. 2,

18 pp. 93–103, 2021, doi: 10.25077/josi.v20.n2.p93-103.2021.

19 [40] RA Saputra, G. Yudoko, and AF Firman, "Proposed Strategy of Heavy Equipment

20 Overhaul: Case Study of a Coal Mining Contractor in Indonesia," vol. 8, no. 4, pp.

21 140–145, 2023.

22 [41] P. Sliwinski and R. Zimroz, “Fuzzy Risk-Based Maintenance Strategy with

23 Safety,” 2022.

24

Page 19 of 19 MEV_author_submission_template