Quality by design (QbD) became an indispensable part of pharmaceutical quality. To increase product quality, understanding how formulation and manufacturing process variables influence product quality is essential. In that holistic context, this work attempted to develop transferable methods to help increase process understanding, control and robustness. For the purpose of these goals a recombinant protein production process with E. coli as a host was used. The product was formed as inclusion bodies (IB). A high throughput method for IB sizing using nano-particle tracking analysis (NTA) was developed and established. The effect of temperature oscillations on IB solubility could not be investigated as planned due to insufficient heating and cooling performance of the reactor setup. A softsensor, previously developed in the working group, has been further developed and was used for real-time estimation of the specific substrate uptake rate qS. Therefore the obligatory real-time biomass estimation was decoupled from the carbon balance and realized using a weighted average approach, enabling biomass estimation robust against substrate accumulation. Subsequently, the sensor was used to establish a closed loop control of qS. Because of the declining physiological capacity of the host to metabolize substrate (qScrit) during induction phase, a logical query was developed to detect reaching qScrit in real-time. The physiological closed loop control was used to successfully avoid substrate and metabolite accumulation throughout induction phase of an industrial relevant production process.