Most of the enterprises recognized the importance of Data governance and started data governance programs either at the Enterprise at the individual Business units level. Data governance councils at various levels in an enterprise define and enforce data quality and security, using policies, standards, and Procedures. The success of Data governance program is heavily relying on people or team who is governing the data governance council. However, Data governance is best implemented by leveraging people, process, and Technology. In the past, subject matter experts of individual data domains maintained the metadata, which is a critical component of data governance. Efficient and automated metadata management, which can be established today by leveraging technology and process not just SMEs, has the potential to mitigate the risk of people dependency. Smart factories, which are heavily automated, generates data at a scale and speed which were never seen before by manufacturing industry, and face the crisis to maintain the data quality, and to make the data available for analytics for any further use. Data generated by smart factories are diverse and are mostly stored in distributed systems, which further increases the complexity of data governance through metadata management. Efficient metadata management, mostly automated, can help smart factories to achieve Data governance goals, and help to provide data as a service. This paper discusses the shortcomings of Hybrid data governance model – acknowledged as the better model for data governance by industry – and proposes a system architecture which has all the benefits of the hybrid model in addition to improvements that are necessary for a smart factory.