Applied reliability methodologies for the development of medical devices in the pharma industry

Applied reliability methodologies for the development of medical devices in the pharma industry. The reliability engineering field is vast and multisectoral, and it can be described as the intersection of different scientific areas such as mathematics, statistics, engineering, and data science. Nowadays, companies are integrating new technologies such as cloud computing, artificial intelligence, and machine learning. This process is called the 4th industrial revolution, and reliability engineering is a key enabler of this transformation by addressing the challenges associated with implementing advanced technologies. This thesis project aims to develop an adequate environment to deal with the most important topics related to reliability engineering to obtain a scientifically robust, versatile, and free tool. Sanofi (the company where this thesis work has been conducted) has put effort into testing its products but still needs to step forward in implementing reliability measures during product development. The purpose of the thesis project was to provide a free software that can help with essential reliability tasks and push the company to implement a reliability program. To provide a free software, the chosen platform for the development was RStudio, which is a multifunctional open-source software that is easily interfaceable with Python. The development of the tool was divided into two big sections depending on the type of item studied, which are repairable and non-repairable. This led to further division based on the data that could be censored or non-censored. The reliability approaches have been studied and translated into an RStudio script for each category. In addition to the basic analysis of the data, a section regarding Accelerated Life Tests (ALT) has been implemented because of their importance in cost and time-saving. To develop the software, the limited libraries for reliability engineering of RStudio were enriched with new routines and some Python functions. One of the project's challenges was gathering Sanofi’s data since the data needed to apply a good reliability program were not found. Data from the complaint center was used to test the software, but because of the uncertainty related to data, the results couldn’t give an actual prediction about the items' behavior. This happened because the data needed should come from proper tests until the failure of the items. In contrast, the data used were obtained as the difference between the date of manufacture and the date of customer complaint. The data-related problems were shown, and the company was provided with examples of useful data needed. Because of the increasing amount of data caused by the digital transformation of companies and advanced technologies, organized and available data are the roots of improving efficiency, managing risks, and promoting the company’s growth. However, the data research that was conducted showed an important deficiency in data organization and availability, which is common in multiple companies. To proceed with further analysis of the costs and benefits of the reliability program, the company needs to start proper testing of the items at the early stage of development to detect the design deficiencies in advance and correct them before manufacturing the complete product.