In the cause of increasing automation, modern production systems are highly tuneable. Simultaneously, todays globalizedmarket requires an optimal, flexible and individual production planning, which further complicates the parametrization ofa production process. At the same time new possibilities of facing those challenges arise. In the context of a Smart Factory,a comprehensive and well-organized database provides the knowledge to respond dynamically to new environmentalconditions and tune the process with respect to concurrent gains (e. g. cycle time and plant peak power of the plant).We propose a multi-criteria process optimization based on a digital twin of a production plant. It has been applied to andevaluated on a model factory, which is used for research studies, by means of three demonstrating scenarios: Optimizingenergy consumption within fastest cycle time, limit peak power and reduce wear of a robot.