The use of Internet of Things (IoT) and Artificial Intelligence (AI) is changing the face of industrial automation through the use of predictive analysis and data analysis. The following paper proposes a new approach that integrates IoT and AI to improve prediction and control and improve the performance of industries. Some of the insights that have been established from the study are the design of a flexible architecture that can support the integration of the real-time IoT data with the AI analytics to support decision making. Our framework uses IoT sensors to gather large amounts of operational data that are analyzed using machine learning algorithms to identify possible system failures and inefficiencies. Through the use of predictive maintenance, which is informed by AI, industrial systems are able to reduce on the downtimes and increase on the production. To assess the efficiency of the proposed framework, we performed numerous tests in a realistically modeled industrial setting. Studies show that our model boosts decision-making precision by 25% and cuts system failure by 30% as opposed to conventional rule-based systems. Furthermore, the AI algorithms used in the framework were generally applicable across the industrial domains and hence the ability of the framework can be applied across different industries such as manufacturing and energy management. This work shows that IoT and AI integration in industry has the capability to enhance the industrial automation and can help to achieve a significant enhancement in the areas of predictive maintenance, operation, and decision making. Future work will involve the fine-tuning of the model for practical use, the investigation of new machine learning algorithms for higher accuracy of the prediction and the expansion of the framework to cover additional large-scale industrial systems