|Abstract||Cane sugar is manufactured in process vessels called vacuum pans, which are closely supervised by operating personnel, known as pan boilers. Traditionally, the pan boiler’s job has been made easier through the use of a control system based on electrical conductivity. However, conductivity-based vacuum pan control is far from perfect, since it requires continual attention from the pan boiler to ensure that the vacuum pan is running efficiently. Existing and emerging technologies are now available which should allow for more advanced monitoring and control schemes to be developed for vacuum pans. The ultimate goal for these technologies is to develop a control system that can run the pan closer to its optimal operating point, with less operator intervention. The main aim of this project is to develop optimal operating policies for an industrial vacuum pan and implement them on a production scale. Rather than use conductivity as the control variable, it is proposed that more process relevant variables should be considered. These variables are the molasses sucrose oversaturation, or OS, and the massecuite crystal content, or CC. Once these policies have been developed, they need to be implemented. Before that can happen, the OS and CC variables need to be available to the controller. The problem is that the variables cannot be measured directly – there are no devices that can measure either of these variables. However, a technique called soft sensing can be used to estimate the values of OS and CC in the vacuum pan. A soft sensor is actually a computer model of the vacuum pan that uses other available process measurements to predict the values for the OS and CC in the pan. This project’s second aim is to develop a soft sensor that can be implemented on a factory vacuum pan, so that the OS and CC variables are available for process control purposes. The third aim of the project is to evaluate the effectiveness of controlling the vacuum pan by using OS and CC, in place of electrical conductivity. The final aim of this project is to control the full-scale vacuum pan in such a way that it follows the predicted optimal operating policies. This final aim is very important to the Australian sugar industry, since it can help to make each vacuum pan operate faster and therefore be more productive. An extension of this goal is that the vacuum pan might also run with less operator intervention, giving them more time to see to other, more pressing tasks in the factory.