"A competitive weapon for companies in process-intensive industries such as the chemical industry, is effective mastery of supply chain management from raw material to end product so as to improve efficiency, effectiveness, & profitability. However, in flow scheduling, the human operator can be overwhelmed by the complexity of the problem. CAP-SCHED involves developing a new intelligent scheduling system for companies in continuous/semi-continuous process industries. The innovative solution uses a multi-agent framework to combine 2 existing approaches which, individually, are inadequate to improve scheduling. However, combining both approaches offers significant improvements to implement more profitable solutions to scheduling problems, with better account taken of constraints. The 2 approaches are: 1. Simulation through trial & error: * The scheduler instructs the system to calculate all material flows in terms of quantity & quality from up- to down-stream throughout the plant; * If infeasibility occurs (e.g. capacity limitation…) then the planner readjusts parameters in trial & error mode to reach a feasible solution. and 2. Algorithmic constraint solving: * the problem is represented as a mathematical model & the system attempts a solution, helped by heuristic guesses so as to guide the search algorithm via domain knowledge. The new idea is to allow full interaction with the system via a Graphical User Interface as if running in simulation mode, & to translate instructions, when the value of an operating parameter (or the starting date of an event…) is fixed, in terms of constraints imposed in the mathematical model. Thus, the person responsible for scheduling can choose to fix variable values, and to let the system find those left free so as to reach a feasible solution. Overall, the solution will drastically reduce the need for multiple planning/scheduling loops so as to achieve a consistently optimised solution with improved profitability."