A Standardised Framework for Data Science in Advanced Manufacturing Systems
Year: 2024
Type: article
Abstract: As the field of Data Science has been more widely adopted by organisations in recent years, the idea of defined workflows for efficient application has become a popular topic for discussion. Frameworks such as the Microsoft Team Data Science workflow, Cross-industry standard process for data mining (CRISP-DM), and community-led project templates through Cookiecutter have looked to standardise and optimise the way in which Data Science projects are conducted from end-to-end. Due to the extensive range of applications in advanced manufacturing systems, researchers engaged in the area are, understandably, more-often specialists in their technical niche than they are in the technical delivery of data science projects; though, there is a strong argument to be made that many of their activities fall under the banner of data science in some shape or form. To date, there has not been a published example of a comprehensive framework which captures the requirements of a rigorous data science workflow, whilst also specifically tackling the distinct technical challenges present in the specialised areas of advanced manufacturing, such as machine tool connectivity, machining trial planning or manufactured component quality assurance. In this paper, the popular, published frameworks are reviewed, and their suitability are assessed for a use case in advanced manufacturing research. The popular frameworks are analysed for their similarities and differences, as well as the potential missing components when applied to a typical advanced manufacturing workflow. This work proposes an extended framework, which effectively meets the needs for research activity in advanced manufacturing systems research to improve their effectiveness, applicability, rigour, and efficiency of delivery.
Source: Procedia CIRP
Authors: Tim Rooker, Erica Smith, Javier Dominguez-Caballero, Lindsay Lee, Sabino Ayvar-Soberanis
Language: English
Subfield: Industrial and Manufacturing Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4404789484