How Big Data Is Changing Manufacturing


Manufacturing organizations are faced with many challenges to bring new innovative products and services to market in short time and also to improve overall efficiencies. Industrial IOT or Industry 4.0 brings together many technology solutions for manufacturing organization to meet these challenges.

Engineering products are becoming smarter, connected, intelligent, self-learning and autonomous. The design and development philosophy is completely different for these smart and connected products compared to the conventional products. During the design of these smart products, all the operational and maintenance parameters to be captured have to be identified. Sensors, data acquisition and communication systems have to be designed and integrated with the structure or system. Similarly, manufacturing and testing of these smart products is radically different from the conventional products. Further, many software components and algorithms need to be integrated with the system to make it more intelligent. 

Design of smart and connected products require a systemic perspective on various systems and sub-systems of the product, how they need to be integrated and operated to meet specific functional requirements. Specific failure modes of the structure and system needs to be identified before a product is realized. This requires lot of domain expertise of the system being designed. Many technology aspects like data interoperability, data standards and security have to be considered. Knowledge Based Engineering (KBE) can help in capturing the product and process knowledge. This knowledge can be effectively used to identify failure modes and parameters to be captured for monitoring the performance of the system.

One of the biggest advantages of Big Data technology is that we are in a position to compare the performance of a product real time now. The operational data need to be verified against specific threshold to check whether it is working fine or not. The parameter thresholds are set based on the domain expertise. In case these thresholds are not available, machine learning techniques can be utilized on the historical data to identify the operating range and thresholds. In absence of Big Data technologies, the operating parameters are used to raise alarms only. However, now the operating parameters data can be used to do rigorous trend analysis which can then be utilized for prognostics mainly for remaining useful life computation.  This will require development of advanced hybrid models bringing together best of data sciences and physics. The data science models can be statistical models or advanced machine learning techniques.

In summary, Big Data based advanced models can help in improving the efficiencies across the manufacturing enterprise. These include:

Engineering efficiency – This will be related to reducing the cycle time required to design, analyze, optimize and certify a product.

- Supply chain efficiency – Engineering organizations need to interact with supply chain eco-system for raw materials, components and sub-systems. Currently organizations face many challenges of supplier quality and efficiency. 

Manufacturing shop floor efficiencies

o   Operational – This is related to operational efficiency of a manufacturing shop floor which is a combined metric of performance, quality and availability. Organizations struggle to keep the Operational Equipment Effectiveness (OEE) more than 60% currently. Every one percent improvement in OEE will have direct influence on companies top and bottom lines

o   Maintenance – It is essential to reduce the maintenance cost of equipment in the manufacturing shop floor and keep the assets available all the time. Organizations need to classify their assets based on their criticality and critical assets may have to go for condition based monitoring. Back bone of condition based monitoring is Big Data technology

o   Information – Information efficiency is back bone for a manufacturing organization. It is essential that data goes smoothly from shop floor to top floor. Diverse and complex equipment will pose many information challenges to organizations such as data interoperability, security etc. Big Data based platforms can help in improving the information efficiency 

o   Energy – Organizations need to reduce their wastage and also reduce the energy consumption. This will be part of the sustainability initiative within the organization and across the eco-system

Service Efficiency – Once products goes out of the manufacturing shop floor and are with the end customers it is essential that they meet the end customer expectation in terms of its operational cost or maintenance cost. OEMs are no more product sellers and are entering into long term service agreements which are creating new business models. This necessitates the need to improve the service efficiency of end products