Development of model-based methods to streamline Research and Development in the Pharmaceutical Industry

The pharmaceutical industry has contributed positively to global longevity and well-being, and as well to the economy. However, several challenges must be faced by the pharmaceutical R&D, like the increasing difficulty in developing new treatments with significant advantages over the existing ones and the need to accelerate the entire R&D process, which often lasts more than half of the patent life and requires substantial investments of resources, labor and money. In this Dissertation, novel model-based methods are developed in order to streamline pharmaceutical R&D, while improving product and process understanding and ensuring product quality and process robustness. Specifically, the methods proposed in this Dissertation allow to: (i) streamline the design of tablets lubrication; (ii) minimise model prediction variance in the whole design space by means of a trade-off between space exploration and information maximisation, (iii) adapt the novel procedure for the minimisation of model prediction variance to the operation of an automated platform in order to streamline kinetic studies; (iv) implement a fully autonomous operation of a chemical platform with the aim of collecting experiments to estimate model parameters and minimise model prediction variance; (v) accurately predict drug solubility in mixtures of organic solvents; (vi) predict drug solubility in intestinal fluids, considering the effects of food and physiological factors.