To achieve the sustainable operation of machine tools, it is essential to address the optimization problem consisting of productivity, quality, and energy use. The thermal behavior significantly influences these parameters, thereby directly affecting sustainable machine operation. The Collaborative Research Center CRC/TRR96, funded by the German Research Foundation DFG, has proposed approaches to systematically tackle this conflict. Building on the foundational knowledge and transferring it into practical applications, control systems can effectively be developed and implemented. One such application is the model predictive control of active cooling systems in machine tools for compensation of thermoelastic displacements during operation, thus enhancing the machine accuracy. However, to optimally select the operating point of the active cooling system from the holistic sustainability perspective, it is crucial to consider the effects on all dimensions of the optimization problem. The design of a multivariable model predictive control is demonstrated using the example of the spindle cooling. The approach involves extending the system modeling and developing of a sustainability index.