Optimising Skill Matching in the Service Industry for Large Multi-skilled Workforces

The continued drive to improve efficiency within the service operations sector is motivating the development of more sophisticated service chain planning tools to aid in longer term planning decisions. This involves optimising resource against expected demand and is critical for successful operations of service industries with large multi-skilled workforces, such as telecoms, utility companies and logistic companies. To effectively plan over longer durations a key requirement is the ability to simulate the effects any long term decisions have on the shorter term planning processes. For this purpose, a mathematical model encapsulating all the factors of the shorter term planning, such as skills, geographical constraints, and other business objectives was defined. Attempting to use conventional methods to optimise over this model highlighted poor scalability as the complexity increased. This has motivated the development of a heuristic method to provide near optimal solutions to the model in a shorter timescale. The specific problem we look at is that of matching resource to demand across the skill dimension. We design a genetic algorithm to solve this problem and show that it produces better solutions than a current planning approach, providing a powerful means to automate that process. We also show it reaching near optimal solutions in all cases, proving it is a feasible replacement for the poorly scaling linear model approach.