Skip to main content

Fine-tuning Production in Manufacturing with Big Data Analytics

According to a study, 1.7 MB of average data volume was created per second in 2020 and the total size of data on the Cloud is estimated to be 44 zettabytes, (a trillion gigabytes), by the end of 2025[1]. The data revolution has impacted all forms of businesses, including manufacturing. Moreover, the manufacturing sector has been using several software such as Enterprise Resource Planning (ERP) systems and quality assurance systems for years, without fully tapping into the potential of using historical data that these systems have accumulated over time.

With the advent of state-of-the-art technologies such as Internet of Things (IoT), data type, volume and the rate of data generation have scaled up. As a result, the need for technologies that can handle, integrate, store and process data from different types of sources and process it in real time, has further intensified. The manufacturing sector is swiftly moving into the era of Industry 4.0 where the vision for smart factories will completely come into effect. This would mean extensive sensor deployments, use of sophisticated data communication paradigms and development of intelligent solutions that can bestow high degree of autonomy on systems. Big data analytics has a key role to play in bringing all these different facets together to provide actionable insights.

The implications of using big data technologies in manufacturing are multi-faceted, from effective data acquisition, heterogeneous device data management to reactive decision making and predictive/prescriptive solutions. Recent past has witnessed rapid expansion in technology adoption for the manufacturing sector, which is driven by efforts for sustainability and gaining competitive advantage. Most of these efforts are focused on identification of the right production pipeline, from the business perspective. Big data can support a plethora of applications for the manufacturing sector, which can broadly be classified into the following three areas:

  • Improving operational efficiency of manufacturing
  • Supporting quality assurance
  • Enhancing supply chain efficiency

Use cases of big data analytics for manufacturing are summarized in Figure 1. Most of the applications stem from the factory floor tasks that require human intervention. Therefore, these include actions requiring adjustments that need to be made for maximizing throughput, optimizing production or eliminating the occurrence of defects or failures. In addition, optimizing energy consumption and boosting efficiency of equipment also include tasks that can be modelled using data-driven systems.

Figure 1 – Use Cases of Big Data Analytics in Manufacturing

Applications of Big Data Analytics in Manufacturing Sector

Big data technology can support a range of systems and applications in the manufacturing sector. Some of these system-level applications are discussed in the following sections.

System-Supported Decision Making

From a technical perspective, some of the data-driven, big data solutions such as decision support systems and recommendation systems can smoothly fit into several application-level use cases[2]. Decision Support Systems (DSS) gather and process data to provide useful insights for effective decision making. Moreover, they may be partially or completely autonomous.  As part of Industry 5.0[3] revolution, there will be a development of hybrid DSSs that are autonomous for repetitive, non-critical tasks which will require human monitoring for vital decision making such as time of task initiation. Big data recommendation engines support human action-based interventions by suggesting optimal choices. 

Demand Prediction and Tailored Responses to Market Fluctuations

The most remarkable offering of big data technologies is the power to predict the future, even in real time for most cases. One of the simplest use cases that demonstrate this capability in action is a system that analyses data from Customer Relationship Management (CRM) systems to identify patterns in orders and product consumption[4]. Insights from such systems drive adjustments in the manufacturing pipeline to ensure production is enough to support incoming orders and the inventory is ready before a sales cycle reaches completion. This in turn ensures optimum production hence limiting wastage.

Fault Prevention and Quality Assurance Systems

In order to reduce waste, it is vital to identify in-process deviations. Big data technologies can play a key role in acquiring and processing data from sensors deployed on tool jaws and shafts to predict any possible deviations that may have been introduced. The system can be set up to send an alert notification when the deviation values cross a preset threshold or when quality assurance tolerance levels are not met, hence reducing the amount of products that might eventually be rejected.[5].

Systems for Preventive and Predictive Maintenance

As mentioned above, big data technologies can tap into data collected from different sensor arrays to make specific predictions. Within the context of equipment, similar analyses can predict equipment failure with prescription for possible interventions. There are several advantages of using this approach[6]. Firstly, preventive and predictive maintenance systems improve availability by reducing downtime. Besides this, such systems can play a vital role in preventing irreversible damages to equipment, prolonging equipment lifespan and saving costs.

Choosing Big Data Platform for your Manufacturing Business

The vision of every manufacturing business is centred on achieving perfect production. The standard metric for evaluating the efficacy of manufacturing processes is Overall Equipment Effectiveness (OEE)[7]. In view of the inefficiencies associated with manufacturing pipelines, the globally acceptable threshold for OEE is set at 85%[8]. This demonstrates how far the production processes are from perfection and any percentage change in this value will not just improve profitability, but it will also mean competitive advantage. Effective use of data can play an instrumental role in boosting productivity, in this regard.  

There is immense scope of innovation in this field of research. Many novel applications can be developed to support existing systems and improve the overall efficiency and productivity of this industry. Collaborative efforts between experts from manufacturing industry and technologists is sure to take modernization and adoption of technology in the manufacturing industry to a higher level of maturity.  

How can we help?

If you have any questions about big data adoption for your business or would like to discuss your specific requirements, you can email the University of Wolverhampton SOLVD team solvd@wlc.ac.uk or visit www.wlv.ac.uk/solvd. The SOLVD project supports Telford & Wrekin and Shropshire businesses with the adoption of digital technologies to improve productivity and growth. Eligible businesses can access 12 hours of fully funded support with academic experts.

Blog by Dr Samiya Khan - Research Fellow in Big Data, Artificial Intelligence & Edge Computing at The University of Wolverhampton.

 

[1] https://techjury.net/blog/how-much-data-is-created-every-day/#gref

[2] https://link.springer.com/article/10.1007/s00170-020-06389-1

[3] https://www.i-scoop.eu/industry-4-0/industry-5-0/

[4] https://www.altexsoft.com/blog/demand-forecasting-methods-using-machine-learning/

[5] https://www.tandfonline.com/doi/full/10.1080/0951192X.2020.1718765

[6] https://techhq.com/2021/01/how-predictive-maintenance-is-changing-the-industrial-enterprise-for-good/

[7] https://www.oee.com/

[8] https://www.oee.com/world-class-oee.html