Today, data science presents immense opportunities by turning raw data into manufacturing intelligence in data-driven manufacturing that aims to improve operational efficiency and product quality together with reducing costs and risks. However, manufacturing firms face difficulties in managing their data science endeavors for reaping these potential benefits. Maturity models are developed to guide organizations by providing an extensive roadmap for improvement in certain areas. Therefore, this paper seeks to address this problem by proposing a theoretically grounded Data Science Maturity Model (DSMM) for manufacturing organizations to assess their existing strengths and weaknesses, perform a gap analysis, and draw a roadmap for continuous improvements in their progress towards data-driven manufacturing. DSMM comprises six maturity levels from “Not Performed” to” Innovating” and twenty-eight data science processes categorized under six headings: Organization, Strategy Management, Data Analytics, Data Governance, Technology Management, and Supporting. The applicability and usefulness of DSMM are validated through multiple case studies conducted in manufacturing organizations of various sizes, industries, and countries. The case study results indicate that DSMM is applicable in different settings and is able to reflect the organizations’ current data science maturity levels and provide significant insights to improve their data science capabilities.

今天，数据科学通过将原始数据转化为数据驱动制造中的制造智能提供了巨大的机会，旨在提高运营效率和产品质量，同时降低成本和风险。然而，制造公司在管理他们的数据科学努力以获取这些潜在利益方面面临困难。成熟度模型旨在通过为某些领域的改进提供广泛的路线图来指导组织。因此，本文试图通过为制造组织提出一个基于理论的数据科学成熟度模型 (DSMM) 来解决这个问题，以评估他们现有的优势和劣势，进行差距分析，并绘制路线图，以不断改进他们在数据方面的进步。驱动制造。DSMM 包括从“未执行”到“创新”的六个成熟度级别和 28 个数据科学过程，分为六个标题：组织、战略管理、数据分析、数据治理、技术管理和支持。DSMM 的适用性和有用性通过在不同规模、行业和国家的制造组织中进行的多个案例研究得到验证。案例研究结果表明，DSMM 适用于不同的环境，能够反映组织当前的数据科学成熟度水平，并为提高其数据科学能力提供重要见解。和支持。DSMM 的适用性和有用性通过在不同规模、行业和国家的制造组织中进行的多个案例研究得到验证。案例研究结果表明，DSMM 适用于不同的环境，能够反映组织当前的数据科学成熟度水平，并为提高其数据科学能力提供重要见解。和支持。DSMM 的适用性和有用性通过在不同规模、行业和国家的制造组织中进行的多个案例研究得到验证。案例研究结果表明，DSMM 适用于不同的环境，能够反映组织当前的数据科学成熟度水平，并为提高其数据科学能力提供重要见解。