Digital Twin Framework for Real-Time Monitoring and Optimization of Ultrasonic Vibration-Assisted EDM using AI and IoT

Tanvi Muttin; Sunayana P Singh; Tameem Ulla; Vagarth Pandey

doi:10.70454/IJMRE.2026.60104

Authors

Tanvi Muttin RV Institute Of Technology And Management Author
Sunayana P Singh RV Institute Of Technology and Management Author
Tameem Ulla RV Institute of Technology and Management Author
Vagarth Pandey RV Institute of Technology and Management Author

DOI:

https://doi.org/10.70454/IJMRE.2026.60104

Keywords:

Digital Twin, UV-EDM, Gaussian Process Regression, NSGA-II, Analytic Hierarchy Process, IoT, Multi Objective Optimization

Abstract

Ultrasonic vibration-assisted electrical discharge machining (UV-EDM) requires careful tuning of multiple in teracting parameters to balance Material Removal Rate (MRR) and surface roughness (Ra). This paper presents an integrated Digital Twin framework coupling real-time IoT sensor streams with Gaussian Process Regression (GPR) surrogate modeling, NSGA-II multi-objective optimization, and Analytic Hierarchy Process (AHP) decision support. Validation on 150 experimental UV-EDM trials demonstrates: (1) predictive R2 = 0.92 for MRR and R2 = 0.89 for Ra, (2) a 14.6% improvement in MRR, and (3) a 9.2% reduction in Ra compared to baseline settings. Beyond numerical performance, the proposed system high lights the value of hybrid intelligence by combining data-driven learning with physical domain understanding to achieve con sistent machining quality under varying operational conditions. The digital twin’s continuous feedback loop allows operators to visualize tool wear, spark energy, and machining stability in real time, promoting proactive decision-making rather than reactive control. Furthermore, the framework’s modular design ensures scalability across different EDM configurations, mak ing it adaptable for both research and industrial deployment. Overall, this work establishes a practical pathway toward smart manufacturing by bridging the gap between physical machining and its virtual counterpart through AI and IoT integration.

References

[1]. V.-T. Dinh, T.-Q. Le, T.-T. Do, N.-P. Vu, and T.-P. Tran, “Multi-objective optimization of material removal rate and surface roughness in ultrasonic vibration-assisted EDM using NSGA-II, GPR, and AHP,” Engineering, Technology & Applied Science Research, vol. 13, no. 1, pp. 9876–9889, 2025.

[2]. V.-T. Dinh, T.-Q. Le, D.-B. Vu, N.-P. Vu, and T.-L. Mai, “Ultrasonic EDMfor external cylindrical surface machining: Horn design and hybrid NSGA-II–AHP optimization,” Machines, vol. 13, no. 2, p. 89, 2025.

[3]. C. E. Rasmussen and C. K. I. Williams, Gaussian Processes for Machine Learning. MIT Press, 2006.

[4]. K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan, “A fast and elitist multiobjective genetic algorithm: NSGA-II,” IEEE Transactions on Evolutionary Computation, vol. 6, no. 2, pp. 182–197, 2002.

[5]. T. L. Saaty, The Analytic Hierarchy Process: Planning, Priority Setting, Resource Allocation. McGraw-Hill, 1980

[6]. X. Fu, Y. Zhang, B. Shi, and L. Wang, “Digital twin technology in modern machining: Architectures, applications, and future perspectives,” Journal of Manufacturing Systems, vol. 75, pp. 1–15, 2025.

[7]. G. Pronost, “Digital twins along the product lifecycle,” Digital Twin Journal, vol. 2, no. 3, pp. 45–62, 2024.

[8]. Y. Wang, Z. Liu, M. Chen, and X. Zhou, “Acoustic sensors for monitor ing partial discharge and anomalies in electrical discharge machining,” Sensors, vol. 24, no. 8, p. 2467, 2024.

[9]. S. Natarajan, K. Sharma, and R. Panda, “Digital twin-driven tool condition monitoring and remaining useful life prediction in CNC machining,” Journal of Intelligent Manufacturing, vol. 34, no. 7, pp. 2891–2908, 2023.

[10]. OASIS Open Standards, “MQTT Version 5.0,” Standard document, 2019. [Online]. Available: https://docs.oasis-open.org/mqtt/mqtt/v5.0/

[11]. OPC Foundation, “OPC UA — Specifications and Pro f iles,” IEC/OPC UA Standard, 2023. [Online]. Available: https://reference.opcfoundation.org/

[12]. B. Chen et al., “Effect of ultrasonic-assisted EDM on material removal rate, electrode wear, and surface quality of titanium alloy,” Scientific Reports, vol. 14, p. 8234, 2024.

[13]. S. Guo et al., “Comparative study of ultrasonic-assisted die-sinking micro-EDM: Mechanisms and performance improvements,” Sensors, vol. 24, no. 10, p. 3142, 2024.

[14]. C. Li, D. Huang, R. Wang, Z. Liu, and Y. Jin, “Safe active learning for multi-output Gaussian processes: Application to urban environmental monitoring,” Proceedings of Machine Learning Research, vol. 108, pp. 234–248, 2022.

[15]. F. Pedregosa et al., “Scikit-learn: Machine learning in Python,” Journal of Machine Learning Research, vol. 12, pp. 2825–2830, 2011.

[16]. F.-A. Fortin, F.-M. De Rainville, M.-A. Gardner, M. Parizeau, and C. Gagn´e, “DEAP: Evolutionary algorithms made easy,” Journal of Machine Learning Research, vol. 13, pp. 2171–2175, 2012.

[17]. Plotly Technologies Inc., “Plotly Dash — analytical web applications in Python,” 2020. [Online]. Available: https://dash.plotly.com/

[18]. S. Kerrouchi, B. Lamrabet, and A. Hadj-Tayeb, “Production Digital Twin: A systematic literature review and future research directions,” Journal of Manufacturing Science and Engineering, vol. 146, no. 3, p. 031007, 2024.

[19]. A. Kizhakkinan et al., “Surrogate model development for high-fidelity manufacturing simulations using Gaussian processes and tensor train decomposition,” Journal of Intelligent Manufacturing, vol. 34, no. 6, pp. 2543–2559, 2023.

[20]. M. Matveeva et al., “Application of Gaussian process regression as a sur rogate modeling method for expensive mechanical system simulations,” PMC Biophysics, vol. 5, no. 1, p. 12, 2022