The concept of process analytical technology (pat) has been introduced in pharmaceutical and biopharmaceutical manufacturing over 10 years ago. Yet, the conventional manufacturing paradigm focused on quality by testing (qbt) is still deep-rooted in the quality culture of pharmaceutical companies and, has considerably slowed down the introduction of new technologies in manufacturing. This thesis specifically concerns this issue by focusing on the development of pat applications combined with chemometrics to design, analyse, control and optimize pharmaceutical processes. The challenges underpinning the lack of data-driven decision making in (bio)pharmaceutical manufacturing are discussed in three different perspectives: -! Gathering data and extracting information from pat to build process understanding, fasten product development or introduce improvements in existing processes; -! Setting appropriate workflows for analytical method development and lifecycle management of pat procedures in drug manufacturing, including knowledge transfer and knowledge management in a global environment. -! Establishing knowledge and data-driven approaches where all process unit operations are linked and the product lifecycle perspective is considered under the evolving regulatory framework. The emergence of a data-driven mind set and the advancements in data analytics and computer science are an opportunity for pharmaceutical companies to gain novel insights to improve drug development and manufacturing efficiency. Paper i describes how pat combined with different chemometric approaches can be used to support the conversation of a conventional batch process into a continuous one. The work describes a roadmap for pat screening and details on the scientific basis to develop monitoring and control strategies for continuous reactions based on pat information. Poster i and poster ii further detail on the experimental work required to characterize the reaction system investigated in the study (i. E. , reactants and products). In a very competitive landscape for the launch of biopharmaceutical products, it is crucial to expedite development timelines while maximizing the efficiency of process characterization studies. Paper ii conveys how 2d fluorescence and advanced data analysis methods can shed some light into biologic drug process development. Due to limited sensor capabilities and/or first-principles understanding, 2d fluorescence is a promising technology for accelerating bioprocesses development and evaluating control strategies. When developing a new process or addressing a problem in an existing one, it is fundamental to adopt consistent procedures and practices, particularly in a globalized environment. Perhaps one of the major constraints delaying the adoption of pat in (bio)pharmaceutical manufacturing is the inexistence of a systematic workflow for pat method development and maintenance in routine use. Book chapter i aims to address this gap by proposing a systematic procedure for pat-based methods development and lifecycle management, aligned with current regulatory expectations, and applicable to the production of any (bio)pharmaceutical product. Many of the existing solutions to deal with multivariate data within pat and more general within quality by design (qbd), tend to focus on local data analysis: data is generated and analysed in the context of a microsystem (viz. , sample or unit operation). However, the true benefits of innovative process technologies or advanced data analysis methods can only be realized if the knowledge is properly transferred and maintained throughout the process flowsheet (i. E. , linkage of steps) and over the product lifecycle, including development and commercial manufacturing. Book chapter ii sets continued process verification (cpv) at the center of the process/product lifecycle approach. The cpv concept can only be fully accomplished if the process performs at any point of its existence as well and as consistently as it did when filed and approved in the first place. A workflow to streamline the information hidden in complex databases is provided to elevate legacy product validations to a higher level, in terms of compliance with current regulations, robustness and operational performance. Under the framework of pharmaceutical quality systems (cf. Ich q10, 2009), that incorporate quality risk management (qrm) and data-based justifications to develop a good grasp of all important variability sources, different stakeholders can challenge process owners to make evidence of product quality and consistency which is a true indicator of in-depth process understanding and efficient knowledge dissemination.