Spinning is the textile industry’s lifeblood industry,where the fibre is drawn out,twisted,and wrapped onto a bobbin by using state-of-the-art twisting techniques.Textile intelligent spinning systems are highly dynamic and flexible embedded systems suitable for uncertain environments and physically distributed with sporadic connectivity.The apparent path for success is to ensure the complex textile spinning systems’ proper functionality and transform the existing system simultaneously by adding new artificial intelligence（AI）based predictive modelling techniques that make the manufacturing industry future-ready.With the continuous development in the manufacturing industry and after Industry 4.0,digitalization in the manufacturing sector achieved an exceptional and remarkable milestone.Blistering evolution like cyber-physical production systems（CPS）and digital twin（DT）are gaining popularity due to their considerable impacts on the realization of intelligent manufacturing.It allows a highly controlled mechanism based on monitored algorithms within which data and information from different systems used to increase self-awareness,self-prediction,and self-configuration functionalities.The cyber-physical systems（CPS）and the industrial Internet of things（IIo T）are the key technologies to reach this goal.Based on a CPS,the smart textile industry is a highly sophisticated and integrated factory involving tight coupling between digital twins（physical and computational）components.Spinning production is a high-speed and continuous manufacturing process that suffers from various dynamic disturbances such as machine breakdown,low quality,and job release delay.Real-time task processing is critical to improving production efficiency to satisfy the requirements of dynamic production.Several other factors also affect the quality of yarn regarding machine efficiency,including spinning speed,friction,tension phasing,fibre diameter,strand thickness,and load.Therefore,by applying predictive analytics using machine learning（ML）techniques（algorithms）,both quantitatively and qualitatively,an effort made to identify failure modes and mitigate downtime.Downtime leads to an increased cost per unit and elevated operating costs despite the loss in revenue,and low maintenance strategies can reduce a plant’s overall production capacity between 10%—30%.Two main key factors considered during the implementation of our proposed model for predictive modelling and optimization.The first factor in developing a prognostics and health monitoring system（PHM）for device,system,and system of systems（So S）is to identify critical components and their impact on spinning frame performance.The second factor considered for the drawing workshop is intelligent transportation scheduling of the sliver cans distribution,which is inevitable in these smart spinning CPS to improve production efficiency and increase the return of investment（ROI）ratio.Further,intelligent virtual equivalents were generated during the integration of digital twins with CPS that reflect the physical structures’ actions and offer predictive insights for advanced decision-making.This thesis provides a systematic framework to integrate cyber-physical systems different perspectives within the context of textile spinning shop-floor from device-level（DL）,system-level（SL）,and system of systemslevel（So SL）maintenance applications,i.e.predictive control,modelling and optimization solution.Different case studies have been provided to demonstrate the proposed model’s feasibility for the high-tech spinning industry:i.In the first case study,the author presents a machine learning approach for condition monitoring（CM）on a DL,based on a regularised neural network using automated diagnostics for spinning manufacturing.Digitalization encapsulates the importance of machine CM,which is subjected to predictive analytics for realizing significant improvements in the performance and reliability of rotating equipment,i.e.,spinning.A part of the thesis contributes to finding disturbances in a running system through real-time data sensing and signal processing via the IIo T.Because the controlled sensor network comprises different critical components（device-level）of the same type of machines,a multi-sensor performance assessment and prediction strategy were developed for the system,which worked as intelligent maintenance and diagnostic.ii.In the second case study,a new data-driven prescriptive maintenance and architectural impulse on an SL,based on a regularised neural network using predictive analytics,proposed successfully for ring-spinning.The paradigm shift in computational infrastructures enormously pressured large-scale linear and non-linear automated assembly systems to eliminate and cut down unscheduled downtime and unexpected stoppages.The fundamental process of predictive modelling is PHM,and it is the tool resulting in the development of many algorithms to predict the remaining useful life（RUL）of industrial equipment,hence improving its efficiency.The spinning sensor network designed for the scheduling process comprises different critical components of the same spinning machine frames containing more than thousands of spindles attached to them.Results showed that it operates with a relatively less amount of training data sets but takes advantage of large volumes of data.This integrated system aims to prognosticate abnormalities,disturbances,and failures by providing condition-based monitoring for each component.iii.In the third case study,aiming at the path-planning and decision-making problem,multiAGV have played an increasingly important role in the multi-stage industries,.e.g.,textile spinning.We recast a framework to investigate the improved genetic algorithm on multi-AGV path optimization within spinning drawing frames to solve the complex multi-AGV manoeuvring scheduling decision and path planning problem on So SL.The study reported in this chapter simplifies the scheduling model to meet the draw-out（drawing）workshop’s real-time application requirements.According to the characteristics of decision variables,the model divides into two decision variables: time-independent variables and time-dependent variables.The first step is to use a genetic algorithm to solve the AGV resource allocation problem based on the AGV resource pool strategy and specify the sliver can’s transportation task.The second step is to determine the AGV transportation scheduling problem based on the sliver cans-AGV matching information obtained in the first step.One significant advantage of the presented approach is that the fitness function is calculated based on the machine selection strategy,AGV resource pool strategy,and the process constraints,determining the scheduling sequence of the AGVs to deliver cans.Moreover,it discovered that double-path decision-making constraints minimize the total path distance of all AGVs,and minimizing single-path distances of each AGVs exerted.By using the improved GA,simulation results showed that the entire path distance was shortened.iv.The fourth and last case study pointed at a problem that leads to the high complexity of the production management tasks in the multi-stage spinning industry,i.e.manual handling slivercans have many jobs,and there is a long turnover period in their semi-finished products.A novel heuristic research was conducted that considered mixed-flow shops So SL scheduling problems with AGV distribution and path planning to prevent conflict and dead-lock by optimizing distribution efficiency and improving the automation degree of sliver-cans distribution in a drawout workshop.In this context,a cross-region shared resource pool and an inter-regional independent resource pool,two AGV predictive scheduling strategies were established for the ring-spinning combing process.Besides completion time,AGV utilization rate and unit AGV time were also analyzed with the production line’s bottleneck process.In the optimal computational experiment,results proved that a draw frame equipped with multi-AGV and coordinated scheduling optimization significantly improves the distribution efficiency.Moreover,flow-shop predictive modelling for multi-AGV resources is scarce in the literature,even though this modelling also produces a control mode for each AGV and,if essential,a preventive maintenance plan.