Table 2.

Summary of AI-enabled DTs in smart manufacturing: machinery and equipment level (Section 2.3).

Subfield	AI-Category	Key Methods	Application-Case	Ref.
Condition Monitoring (Section 2.3.2.1)	Supervised Learning	ANN	Prediction of process forces	[133]
		CNN	Prediction of process forces	[134]
		CNN, SVDD	Defect recognition of steel surfaces	[135]
		CNN-DLSTM based transfer learning	Fault detection of rolling bearings	[136]
	Unsupervised Learning	GAN	Prediction of machining vibration signals	[137]
		Dictionary learning, transfer learning	Wave field prediction for damage detection with ultrasonic guided wave	[138]
	Computational Intelligence	Fuzzy inference	Brake CM of an overhead crane	[139]
Predictive Maintenance (Section 2.3.2.2)	Supervised Learning	PGM, MCMC	Prediction of stress-intensity factors and RUL	[140]
		RCM	Prediction of RUL of a drilling machine	[141]
		RF, particle filter	Prediction of tool wear	[142]
		Deep Stacked GRU	Prediction of tool wear	[143]
		LSTM	Equipment utilization prediction	[144]
		LSTM	Tool condition prognostic model	[145]
		LSTM	Estimation of RUL of the machine components	[146]
	Unsupervised Learning	GMM	Tool failure prediction	[147]
		SSAE-PHMM	Prediction of tool wear	[148]
		SSAE, deep transfer learning	Fault prognosis in a car body-side production line	[149]
		GAN, VAE	Generation of a health indicator for PHM of rotating systems	[150]
		CAE	Construction of a health indicator for bearings	[151]
		Distributed k-means	Assessing MAS for collaborative PdM	[152]
	Computational Intelligence	Bayesian network	Mission planning under uncertainty with respect to fatigue cracking	[153]
Dynamics & Control (Section 2.3.2.3)	Supervised Learning	RNN	Prediction of dynamic states in metal cutting	[154]
		ANN	Prediction of resonances frequencies of a thin bulk acoustic wave resonator	[155]
	Computational Intelligence	Gaussian process	Estimation of single-degree-of-freedom dynamic systems	[156]
		Gaussian process	Prediction of the dynamic response	[157]
		GWO	Optimization of motion control system in machine tools	[158]

SVDD: Support Vector Data Description; RCM: Random Coefficient Model; RF: Random Forest; MCMC: Markov Chain Monte Carlo; GRU: Gated Recurrent Units; (D)LSTM: (Deep) Long Short Term Memory; GMM: Gaussian Mixture Model; GAN: Generative Adversarial Network; SSAE-PHMM: Stack Sparse AutoEncoder Parallel Hidden Markov Model; VAE: Variational AutoEncoder; CAE: Convolutional AutoEncoder; MAS: Multi-Agent System; RNN: Recurrent Neural Network; GWO: Grey Wolf Optimization.