10 big data use cases in manufacturing

big data

 

Nowadays the production industry is described with terms such as big data, smart factory, industry 4.0 and Internet of Things (IoT). These terms are all related to the fourth industrial revolution that is characterized by automation and data exchange in manufacturing technologies. The machines, the products itself and even employees can communicate with each other through sensors, barcodes, GPS signals while creating records of each interaction.  

The amount of data to be stored is growing every day, data acquisition is not a problem anymore. The challenge is to make sense of the data, reveal the patterns in it and use them for operational improvements and to support strategic decision making.

Today’s manufacturing organizations have to find a way to handle and process this unprecedented amount of data. Not all of the generated data can supply with useful information, but according to estimates, 33% of all data could be useful when analyzed. Yet only 0.5% of all available data is processed by companies. This means that manufacturers are not using the remaining 32.5% of data that could provide them with valuable business insights and revenue growth.

 

Big data_sunburst chart

 

The manufacturing industry is the industry most affected by big data trends and possibilities due to the nature and amount of data that is produced by it. Most manufacturers are just starting to discover the potentials of using big data tools, but there are already some pioneers within the biggest manufacturers who have provided some big data uses cases to follow.

 

1.  Risk management

There are several many different areas of supply chain management where big data can be of significant help.
Suppliers now have the choice to share their production data with their partners and customers which creates a complete transparency and a highly effective communication channel for both parties. This way the manufacturer can see exactly whether the supplier is delayed with production or just in time, to then adjust all the related processes and avoid waiting times.
Quality data can also be shared the same way, manufacturers can have all the production and product related quality metrics from their suppliers before even receiving the parts.
By having greater visibility into supplier quality levels and other performance metrics, the manufacturer can have a clear visibility on their supplier portfolio and have insightful data in their hands when it comes to supplier contract negotiations.
Having supplier production and quality information available can also provide all data and insight needed for better risk management. Supplier dependencies are quantifiable and allow the manufacturer to make fact-based decisions when it comes to strategic risk management.

 

2.  Build to order configurations

Manufacturing ‘products to order’ became a trend and not just in the automotive industry but in aviation, computer services, and even consumer goods. The build to order (BTO) production approach is a very efficient and profitable business model. But in order to see real growth from it, a well-defined data platform needs to be in place to analyze customer behavior and sales data. The manufacturer needs to have access to all sales data and be able to make precise predictive analytics to foresee order volumes on each possible configuration and adjust their supply chain accordingly.  Also, sales and production data analysis is needed to identify the profitability of each product configuration. This way manufacturers can define their ideal product portfolio to reach their highest possible revenue at a given time.

 

3.  Improve product quality

Product quality maintenance is of top priority for manufacturers. Most of them already have the data needed to significantly improve quality levels and reduce quality-related costs, but just very few of them can connect their data sources in a way that it would provide actionable insights.

Huge savings can be made when using predictive analytics in testing. One single product might require thousands of different quality tests. The number of the tests required can be hugely reduced if pattern recognition and predictive analytics are used to determine the number and type of tests truly needed, instead of performing all tests on all items.

Production line quality can also be significantly enhanced with big data analytics. Sensor data analysis can detect manufacturing defects early, which reduces the time and cost related to adjusting the production processes.  

 

4.  After sales

The costs of warranties and recalls can easily go out of control even due to the smallest mistakes in the production process. With the help of big data, it is possible to either avoid or to foresee warranty or recall issues, potentially saving significant amounts of money. These warranty related costs are more often than not directly related to the quality of the manufacturing processes, therefore smart analytics tool that process production data can have a major impact on both the manufacturing processes and on the quality of goods manufactured.

 

5.  Track daily production

In order to optimize production quality and yield, manufacturers need to have a daily data flow from their production lines in order to see discrepancies and opportunities in real time. This includes sensor data coming from the production machinery and also financial information that is properly linked together with operational data for analysis purposes. Employee data can also be tracked in real-time by allowing a data exchange between employee badges and production line units.

All the data coming from the production line, therefore, creates continuous opportunities for optimization, cost saving, and prevention, as long as the right tools are available for data analysis.

 

6. Data-driven enterprise growth

By using big data, it became possible to quickly compare the performance of different sites and also to pinpoint the reasons for the differences. In addition to internal production and sales data, it is also possible to analyze entire markets, build what-if scenarios and to use predictive models.

Access to these type of insights means that global growth strategy related questions can be answered based on factual data. Questions like where to open a new factory, which company site should be relocated/closed or whether introducing a new product or not are easy to answer once the related data is collected and analyzed.

 

7.  Predictive and preventive maintenance

Thanks to the sophisticated sensor technology that is readily available nowadays, operational data can be collected and analyzed in real-time from almost any kind of machinery or consumer product.

When operational data is analyzed with pattern recognition method, upcoming failures and need for maintenance can be predicted well in advance. That allows preventing downtimes and costs related to maintenance. In the same time, preventive maintenance will drastically prolong the lifespan of machines by preventing irreversible failures.

Predictive maintenance is a phenomenon that is not only used for industrial but also for consumer products, very often the need for maintenance will depend on the usage of the product. In consumer electronics, producers often track consumer activity on the device to then notify in advance about optimal time of maintenance. This creates an ideal user experience and at the same time drastically reduces maintenance and warranty costs for the manufacturer.

 

8. Overhead tracking

The overhead costs are determining the profitability of each manufacturer. To have real control and visibility over these costs, big data environments are needed with connected data sources and advanced analytics capabilities.
Part standardization is one of the big areas that can hugely contribute to reducing supplier-related costs. It can reach a significant reduction in the part and supplier proliferation. This saves not only costs but time on managing parts data.
Labor cost tracking is another big area that can affect overheads. On average 30-40% of overhead in manufacturing is formed by labor costs. Therefore, it is critical to link not only job roles and wages to certain processes but individuals. Employee badges can be tracked with sensors placed on the shop floor. This way manufacturers can identify the exact cost of each task in a process, broken down to individuals.

 

9. Testing and simulation of new manufacturing processes

The day has come in manufacturing when no risk has to be taken when implementing a new product or process. Both manufacturing processes and the products can be tested before production/implementation.  This is possible thanks to digital twins, virtual reality environments and manufacturing process simulations. Using such environments and tools can allow manufacturers to eliminate the risk from decision-making processes. The aim of the so-called digital transformation of manufacturing companies is to implement such data platforms that make strategic decision making a science.  

 

10.  Logistics

In logistics, the usage of big data is less widespread than in other manufacturing areas. Warehousing and transportation are both areas where big data tools can be used with great Return on Investment, but still, there are only a few companies around the world who is operating data-driven logistic services. Pioneers in the automation of warehousing are DHL, Amazon, and Ocado, to mention just a few examples. These companies substituted most of the human labor force with intelligent robots who systematically move around in the warehouses to pick and collect each individual delivery item. The speed of delivery at these companies has exponentially grown and the costs of human labor and human mistakes as well has been disappearing.

Transportation is also being revolutionized at some of the biggest logistics and manufacturing companies. Big data infrastructures allow them to track freights and weather and road conditions in real time. With that, trucks can be diverted any time on their way when a faster and/or more cost-effective route is possible due to any changes. Natural catastrophes and other unpredictable events can be avoided saving millions of dollars a year for some of the biggest transporters.

 

About Actify

Actify has been supporting manufacturing companies for over 20 years. We help companies in their journey to become a connected enterprise where the data sources are connected and everyone in the company has access to the right data, in time for improved efficiencies.
Our SpinFire and Centro products are both invaluable tools for discrete manufacturers in harnessing the opportunities that their big data environment provides. We support manufacturers in building the infrastructure that they need to collect, connect and manage product and production-related data and information. By linking together originally siloed data sources across the enterprise and through different company sites, we create an innovation platform that allows uninterrupted data flow, data visibility, and data analysis.

 

Watch a video about business intelligence

Listen to Scott Millar and Mark Anderson talking about Product Data Intelligence.
Watch

Please complete the form to request more information.