Recent advancements in sensor technology and storage devices have enabled chemical processes to collect large amounts of operational data using hardware sensors. However, the current sensor technology is still not mature enough to measure all the critical quantities in chemical processes, leading to unstable, inefficient, and unsafe operations. Software (soft) sensors, which employ advanced modelling, control, and optimization techniques, offer a cost-effective alternative for estimating hard-to-measure critical quantities using the available process data. Recent studies have shown significant economic improvements when soft sensors are integrated with process control technologies. Despite these advantages, challenges such as computational complexity and insufficient investigation of data correlations prevent their adoption in the current chemical industry. This project addresses these issues by developing efficient and reliable soft sensors using artificial neural networks, data-driven modelling, and data correlation techniques. It also establishes a systematic approach for selecting datasets from industrial production plants to enhance process monitoring and control. The developed soft sensors will be integrated with the existing commercially available advanced process technology to ensure safe, reliable, and resource-efficient operations. The theoretical developments of this project will be implemented in a software package released as an open-source project to facilitate collaboration with academia and industry. Demonstration on industrial production plants and a laboratory pilot plant is also planned to showcase the benefits of the developed soft sensors in the real-world environment. An effective dissemination plan ensures that the project's outcomes reach the target audience.