The aim of this project is the design and development of tailored model predictive algorithms for compressor room control, executable on constraint CPUs. A compressor room ecosystem consists of compressors, auxiliary devices, buffer vessels, and complex networks of pipes. Every compressor and the auxiliary device (dryers and air separators) that leave the production area are steered via a so-called local controller that adjusts the required pressure, sets alarms, monitors sensors, and controls the different actuators. In many cases, however, customers require air demand that cannot be provided by a single compressor, while air demand can vary throughout time. In both cases, multiple compressors are typically used in parallel. A group of compressors is termed a compressor room and is typically regulated by a central controller whose main purpose is to regulate the compressors in such a way so that they provide pressurized air in the most energy-efficient way and/or ensure an equal number of running hours for all involved compressors. Lately, new applications are emerging with a strong focus on high-quality pressurized airflows (marginal oil and particle contamination, extremely low humidity, etc.). However, current control algorithms do not have the flexibility to support these new applications. Providing control algorithms that have the flexibility to deal with a wide range of requirements (e.g., flow, pressure, air quality) puts serious constraints on the type of control technologies and requires new control algorithms and control concepts. A significantly improved control algorithm will open new opportunities, and it is expected to provide solutions with a further reduction of the energy consumption in the compressor room. MPC is believed to offer a good answer to the latter challenges.