Summary of the project – IoT in Manufacturing

Our client in this case happens to be one of the leading optical fiber cable manufacturers in South-Asia. The project revolved around large sets of data being produced every minute through sensors embedded in highly sophisticated and automated machines on the factory floor. It is important for the operations team to know at every stage how their machines are performing and how their operations can be handled more efficiently.

Challenges of the client

With a steadfast belief in the power of data; our client entrusted us with the task of expanding the overall visibility of their data. They wanted to unify their scattered data spread across more than 10 data sources into a single Operational Data Store (ODS) in a structured manner to launch themselves into drawing inferences as well as being able to achieve modern machine learning driven aspects such as predictive maintenance, demand forecasting etc.

How we added value

We started with a basic plant visit and then moved forward toward analyzing the production process at various stages to understand the types of machines involved in the process and to understand their function in the final output. This understanding helped deign a proper architecture and approach for extracting, transforming and loading the data in a well structured way into a SQL ODS. This forms the base for the next phases of the project where aspects such as an enterprise data warehouse, analytical reporting (Power BI) and data mining are involved. The solution used in this involves a hybrid approach wherein both on-premise as well as cloud service are used.