Paper deals with an algorithm which allows the automatic selection of the best operating point of biological system. This task is one of the subjects of research in the field of metabolic engineering, which deals with control related issues, in particular, with modelling of biological phenomena, as well as, monitoring of the unstable states of biomass growth. The possibility of using specific biosensors and microfluidic system for monitoring, optimizing and controlling of a bioreactor is presented in this paper. To ensure proper experiment control of the bioreactor, a real-time measurement of parameters at the macroscale level and metabolic activity of microorganism cells at the microscale level are relevant. Therefore, much attention has been paid to the description and modelling of cyclical changes in metabolic states. For the determination of key process parameters, a microcalorimeter for measuring the heat of reaction has been applied. The biosensor provides additional information, which is useful in development of an interface for monitoring the bioreactor by a decomposition of measurements including the scale of process. Finally, the paper discusses the problem of model selection describing the bioprocess at the microscale level.