Integration of Data Analytics and Data Mining for Machine Failure Mitigation and Decision Support in Metal–Mechanical Industry
Abstract
1. Introduction
2. Materials and Methods
2.1. Data Collection
2.2. Proposed Framework
- Data cleaning: The removal of records with values outside the expected range (outliers) and exclusion of inconsistent records that could compromise the predictive model’s accuracy.
- Data transformation (categorical encoding): The “Product ID” column, originally containing the categories {L, M, H}, is encoded to numeric values: L → 0, M → 1, H → 2. This transformation ensured compatibility with the decision tree algorithm adopted in this study—Classification and Regression Trees (CARTs)—which requires numeric inputs.
- Feature selection: Based on insights from the DA phase and using the Minimum Redundancy Maximum Relevance (mRMR) method [21], the most relevant attributes for failure prediction are selected.
- Data resampling: To correct class imbalance between failure and non-failure instances, the Synthetic Minority Over-Sampling Technique (SMOTE) is applied to increase the representation of minority classes.
3. Results
3.1. Data Analytics
- Elevated air temperature (mean 302.57 K vs. 299.97 K under normal operation):The slight 2.6 K increase represents a crucial threshold for the system’s thermal dissipation capacity.
- Reduced rotation speed (mean 1337.96 rpm vs. 1540.26 rpm):The approximately 13% drop significantly impairs the natural ventilation and airflow needed for efficient cooling.
- Increased torque (mean 52.78 Nm vs. 39.63 Nm):The approximately 33% rise corresponds to substantially greater heat generation during machine operation.
- Establish evidence-based safe operating limits;
- Dynamically adapt monitoring thresholds to current operating conditions;
- Automate preventive interventions upon the detection of precursor conditions;
- Optimize cooling systems to increase capacity in response to high-risk parameter combinations.
3.2. Data Mining
- If process temperature exceeds 309.80 K and torque is less than or equal to 62.42, there is a high probability of heat dissipation failure (Rule: IF Tool_Wear ≤ 186.50 AND Rot_Speed ≤ 1378.50 AND Torque ≤ 62.42 AND Proc_Temp > 309.80 THEN class: Heat Dissipation Failure).
- If torque is less than or equal to 53.19 and process temperature is greater than 311.89 K, the failure is also classified as heat dissipation (Rule: IF Proc_Temp > 311.89 AND Torque ≤ 53.19 THEN class: Heat Dissipation Failure).
- Conversely, if torque exceeds 53.19 with process temperature above 311.89 K, the tree classifies the event as a random failure, indicating behavioral instability (Rule: IF Proc_Temp > 311.89 AND Torque > 53.19 THEN class: Random Failure).
- If process temperature is less than or equal to 313.30 K, the condition is classified as a power failure (Rule: IF Rot_Speed > 1378.50 AND Torque ≤ 13.50 AND Proc_Temp ≤ 313.30 THEN class: Power Failure).
- If process temperature exceeds 313.30 K under the same configuration, the condition is classified as no failure (Rule: IF Rot_Speed > 1378.50 AND Torque ≤ 13.50 AND Proc_Temp > 313.30 THEN class: No Failure).
- Additionally, if the rotation speed exceeds 2328 rpm, the failure is classified as a power failure regardless of the torque value (Rule: IF Rot_Speed > 1378.50 AND Torque > 13.50 AND Rot_Speed > 2328.00 THEN class: Power Failure).
4. Discussion
5. Conclusions
Author Contributions
Funding
Data Availability Statement
Acknowledgments
Conflicts of Interest
Abbreviations
CART | Classification and Regression Tree |
DA | Data Analytics |
DM | Data Mining |
DT | Decision Tree |
mRMR | Minimum Redundancy–Maximum Relevance |
SMOTE | Synthetic Minority Over-Sampling Technique |
References
- Jha, A.K.; Agi, M.A.N.; Ngai, E.W.T. A note on big data analytics capability development in supply chain. Decis. Support Syst. 2020, 138, 113382. [Google Scholar] [CrossRef]
- Awan, U.; Shamim, S.; Khan, Z.; Zia, N.U.; Shariq, S.M.; Khan, M.N. Big data analytics capability and decision-making: The role of data-driven insight on circular economy performance. Technol. Forecast. Soc. Change 2021, 168, 120766. [Google Scholar] [CrossRef]
- Tabesh, P. Who’s making the decisions? How managers can harness artificial intelligence and remain in charge. J. Bus. Strategy 2022, 43, 373–380. [Google Scholar] [CrossRef]
- Prashar, A. Production planning and control in industry 4.0 environment: A morphological analysis of literature and research agenda. J. Intell. Manuf. 2022, 34, 2513–2528. [Google Scholar] [CrossRef]
- Büttner, K.; Antons, O.; Arlinghaus, J.C. Applied Machine Learning for Production Planning and Control: Overview and Potentials. IFAC-PapersOnLine 2022, 55, 2629–2634. [Google Scholar] [CrossRef]
- Medeiros, G.H.A.; Cao, Q.; Zanni-Merk, C.; Samet, A. Manufacturing as a Service in Industry 4.0: A Multi-Objective Optimization Approach. In Intelligent Decision Technologies. IDT 2020. Smart Innovation, Systems and Technologies; Czarnowski, I., Howlett, R., Jain, L., Eds.; Springer: Singapore, 2020; Volume 193. [Google Scholar] [CrossRef]
- Cañas, H.; Mula, J.; Campuzano-Bolarín, F.; Poler, R. A conceptual framework for smart production planning and control in Industry 4.0. Comput. Ind. Eng. 2022, 173, 108659. [Google Scholar] [CrossRef]
- Romsdal, A.; Sgarbossa, F.; Rahmani, M.; Oluyisola, O.; Strandhagen, J.O. Smart production planning and control: Do all planning environments need to be smart? IFAC-PapersOnLine 2021, 54, 355–360. [Google Scholar] [CrossRef]
- Oluyisola, O.E.; Bhalla, S.; Sgarbossa, F.; Strandhagen, J.O. Designing and developing smart production planning and control systems in the Industry 4.0 era: A methodology and case study. J. Intell. Manuf. 2022, 33, 311–332. [Google Scholar] [CrossRef]
- Colangelo, E.; Fries, C.; Hinrichsen, T.F.; Szaller, Á.; Nick, G. Maturity Model for AI in Smart Production Planning and Control System. Procedia CIRP 2022, 107, 493–498. [Google Scholar] [CrossRef]
- Padovano, A.; Longo, F.; Nicoletti, L.; Gazzaneo, L.; Chiurco, A.; Talarico, S.A. Prescriptive maintenance system for intelligent production planning and control in a smart cyber-physical production line. Procedia CIRP 2021, 104, 1819–1824. [Google Scholar] [CrossRef]
- Chimunhu, P.; Topal, E.; Ajak, A.D.; Asad, W. A review of machine learning applications for underground mine planning and scheduling. Resour. Policy 2022, 77, 102693. [Google Scholar] [CrossRef]
- Harjunkoski, I.; Ikonen, T.; Mostafaei, H.; Deneke, T.; Heljanko, K. Synergistic and intelligent process optimization: First results and open challenges. Ind. Eng. Chem. Res. 2020, 59, 16684–16694. [Google Scholar] [CrossRef]
- Morariu, C.; Morariu, O.; Răileanu, S.; Borangiu, T. Machine learning for predictive scheduling and resource allocation in large scale manufacturing systems. Comput. Ind. 2020, 120, 103244. [Google Scholar] [CrossRef]
- Sharanya, S.; Venkataraman, R.; Murali, G. Estimation of remaining useful life of bearings using reduced affinity propagated clustering. J. Eng. Sci. Technol. 2021, 16, 3737–3756. [Google Scholar]
- Salem, K.; AbdelGwad, E.; Kouta, H. Predicting forced blower failures using machine learning algorithms and vibration data for effective maintenance strategies. J. Fail. Anal. Prev. 2023, 23, 2191–2203. [Google Scholar] [CrossRef]
- Surantha, N.; Gozali, I.D. Evaluation of the improved extreme learning machine for machine failure multiclass classification. Electronics 2023, 12, 3501. [Google Scholar] [CrossRef]
- Shahin, M.; Chen, F.F.; Hosseinzadeh, A.; Zand, N. Using machine learning and deep learning algorithms for downtime minimization in manufacturing systems: An early failure detection diagnostic service. Int. J. Adv. Manuf. Technol. 2023, 128, 3857–3883. [Google Scholar] [CrossRef]
- Hosseinzadeh, A.; Chen, F.F.; Shahin, M.; Bouzary, H. A predictive maintenance approach in manufacturing systems via AI-based early failure detection. Manuf. Lett. 2023, 35, 1179–1186. [Google Scholar] [CrossRef]
- Matzka, S. Explainable artificial intelligence for predictive maintenance applications. In Proceedings of the 2020 Third International Conference on Artificial Intelligence for Industries (ai4i), Irvine, CA, USA, 21–23 September 2020; IEEE: New York, NY, USA, 2020; pp. 69–74. [Google Scholar] [CrossRef]
- Bezerra, F.E.; Neto, G.C.d.O.; Cervi, G.M.; Mazetto, R.F.; de Faria, A.M.; Vido, M.; Lima, G.A.; de Araújo, S.A.; Sampaio, M.; Amorim, M. Impacts of feature selection on predicting machine failures by machine learning algorithms. Appl. Sci. 2024, 14, 3337. [Google Scholar] [CrossRef]
- Mitchell, T.M.; Mitchell, T.M. Machine Learning; McGraw-Hill: New York, NY, USA, 1997; Volume 1, No. 9. [Google Scholar]
- McHugh, M.L. Interrater reliability: The kappa statistic. Biochem. Medica 2012, 22, 276–282. [Google Scholar] [CrossRef]
- de Araújo, S.A.; de Barros, D.F.; da Silva, E.M.; Cardoso, M.V. Applying computational intelligence techniques to improve the decision making of business game players. Soft Comput. 2019, 23, 8753–8763. [Google Scholar] [CrossRef]
- Ahatsi, E.; Olanrewaju, O.A. Enhancing Humanitarian Supply Chain Resilience: Evaluating Artificial Intelligence and Big Data Analytics in Two Nations. Logistics 2025, 9, 64. [Google Scholar] [CrossRef]
- Rezig, S.; Turki, S.; Chakroun, A.; Rezg, N. Optimal Strategy of Unreliable Flexible Production System Using Information System. Logistics 2024, 8, 62. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Araujo, S.A.d.; Bomfim, S.L.; Boukouvalas, D.T.; Lourenço, S.R.; Ibusuki, U.; Oliveira Neto, G.C.d. Integration of Data Analytics and Data Mining for Machine Failure Mitigation and Decision Support in Metal–Mechanical Industry. Logistics 2025, 9, 109. https://doi.org/10.3390/logistics9030109
Araujo SAd, Bomfim SL, Boukouvalas DT, Lourenço SR, Ibusuki U, Oliveira Neto GCd. Integration of Data Analytics and Data Mining for Machine Failure Mitigation and Decision Support in Metal–Mechanical Industry. Logistics. 2025; 9(3):109. https://doi.org/10.3390/logistics9030109
Chicago/Turabian StyleAraujo, Sidnei Alves de, Silas Luiz Bomfim, Dimitria T. Boukouvalas, Sergio Ricardo Lourenço, Ugo Ibusuki, and Geraldo Cardoso de Oliveira Neto. 2025. "Integration of Data Analytics and Data Mining for Machine Failure Mitigation and Decision Support in Metal–Mechanical Industry" Logistics 9, no. 3: 109. https://doi.org/10.3390/logistics9030109
APA StyleAraujo, S. A. d., Bomfim, S. L., Boukouvalas, D. T., Lourenço, S. R., Ibusuki, U., & Oliveira Neto, G. C. d. (2025). Integration of Data Analytics and Data Mining for Machine Failure Mitigation and Decision Support in Metal–Mechanical Industry. Logistics, 9(3), 109. https://doi.org/10.3390/logistics9030109