Smart Pharmaceutical Manufacturing

SPuMoni_Group.jpg

 

Pharmaceutical industry is currently producing significant amounts of electronic data through manufacturing lines increasingly automated via pervasive sensors and devices. Manufacturing line data sources are heterogeneous with various embedded systems controlling the different processes involved in the production of medicines.


Data Integrity and end-to-end traceability have become a key point to be compliant with the different international regulations and guidelines. As an example, in order to release a medicine batch number, it is necessary to ensure that all the data produced is ALCOA (Attributable, Legible, Contemporaneous, Original and Accurate) compliant. Auditable computerised systems are therefore key on pharma production lines, since the industry is becoming increasingly regulated for product quality and patient health purposes. As systems are continuously generating data in various formats, data must be dynamically analysed to ensure the quality and compliance of the overall process.


The main idea of this project is to systematically assess all data produced by computerised production systems in representative pharma environments: (i) design data quality assessment models based on the Data Quality dimensions agreed by the European Institute for Innovation Through Health Data, including rules derived from regulatory documents; and, (ii) identify behaviour patterns of data probability distributions over time and among the manufacturing sources to identify outliers, i.e. data behavioural patterns which can violate ALCOA premises.
To this end, there will be a semi-autonomous data quality control decision support system aiding pharma manufacturing companies to reduce the effort of analysing compliance data. Finally, a system prototype demonstration in an operational environment (Technology Readiness Level 7) will be evaluated using industry-grade real pharmaceutical manufacturing data sets and streams coupled with best pharma industry practices.

Keywords: Blockchain, Big Data, Data Mining, Data Quality, Data Intensive Computing