Manufacturing industries face challenges in meeting targets with respect to profitability, sustainability, and safety on a daily basis. With the advent of technologies like Internet of Things, artificial intelligence, and hyper-scaler cloud platforms, industries are adopting these new technologies to transform their operations. Digital twins are at the heart of such digital transformations. They are being developed and deployed more often as the technology is becoming mature. Challenges faced by manufacturing industries in adopting digital twins, methodologies, and frameworks for development and deployment of digital twins are described in detail in this chapter. A few real-life examples from power utilities and mineral-processing industries on process optimization and predictive maintenance are presented. Recent developments covering physics-informed neural network models, dynamic root cause identification, and security of digital twin models are briefly discussed. Suggestions for future research on digital twin systems for manufacturing industries are provided.