Making Better Manufacturing Decisions With the Help of Machine Learning

Manufacturing today is more complex than ever before. With the increasing use of new technologies in an Industry 4.0 context, data is coming in from more and more sources. We are creating more and more data every day. In fact, according to a recent article in Forbes there are "2.5 quintillion bytes of data created each day, and that pace is only accelerating with the growth of the Internet of Things (IoT). Over the last two years alone 90 percent of the data in the world was generated."

Now when we look at the amount of data that is created in the context of Manufacturing, the Industrial Internet of Things, and Global Supply Chains, we can be sure that it is not much different. We are creating a lot of data. It is coming from shop floor systems such as Equipment, Machines, Sensors, Test Stations, Data Historian as well as from business applications such as ERP, SCM, HCM, CRM, MES, Quality etc.

Does Having A Lot of Data Help? How Easily Can You Answer These Questions?

• Are there patterns in data that strongly relate to yield loss or defects?
• Is there a correlation between product failures in the field or customer returns & the manufacturing process used?
• What are the top influencing factors for quality, yield, and cycle time?
• Can we predict process deviations and product defects early during manufacturing to minimize scrap & rework?
• Can we trace man, machine, method, and material for defective products and identify similar products and impacted customers (smart recall)?

Answering these questions is not easy - despite all the data that exists!

In order to use existing data effectively and efficiently to drive the right set of decisions and actions, you have to address the fact that typically, relevant data is distributed throughout an enterprise. That is, your data sits in Operational Technology (OT) as well as Information Technology (IT). The data from Machines and Equipment, from Enterprise Applications, and from Embedded Data Management Platforms must be acquired, stored, and analyzed. It is not an easy task to accomplish. It makes your life easier if you can leverage automated and manual upload capabilities to ingest data from sensor enabled equipment, machines, and facilities on the shop floor; or if you can ingest data from transactional applications such as MES, Quality Management, LIMS, ERP, SCM, HCM, and CRM; and utilize embedded Oracle PaaS technologies across Database and Big Data stacks. Running all of this on Oracle Cloud Infrastructure (OCI) supports a manufacturing-aware data lake that can store structured, semi-structured, and unstructured data from a variety of sources and also organize the massive data present in the data lake into 5M categories (manpower, machine, method, material, and management).

Once you gather all the data, it is equally important to contextualize and prepare that data to create a comprehensive snapshot of the manufacturing state at any given point in time, to facilitate machine-learning analysis, and to facilitate comprehensive analysis of the entire manufacturing process.

Then, as data is prepared and organized, data scientists can work with the data in the data lake. The data model will change over time, so it is necessary to apply a comprehensive Model Lifecycle Management accordingly, from creation, training, deployment, performance evaluation, and change. Finding data scientists is another story, though, as they are hard to find. In order to attract them, you need to give them modern tools they love to work with.

Get Insights into patterns and correlations and leverage predictive genealogy and traceability analyis

• Use adaptive intelligence to analyze 5M (manpower, machine, material, method, management) related information from manufacturing operations to understand the impact on key business outcomes, with the top influencing factors and variables in the manufacturing environment and from historical data that have the highest influence on key performance metrics, such as yield, quality, cycle time, scrap, rework, and costs.
• Compare current manufacturing conditions against suspect patterns from historical data analysis to predict potential yield loss and product defects
• Receive alerts for predictions that match specific conditions such as confidence%, product context, etc.
• Manage downstream orchestration by subscribing to REST services for predictive alerts (for example, put job on hold, create quality non-conformance, etc.) and create transactions in other applications
• Use an intuitive graph based navigation, traverse back the entire manufacturing process to identify 5M related information.
• For any window of time period, view all relevant manufacturing events such as machine sensor reading anomalies, alarms/alerts, quality test results, work order start/stop, and status changes such as released, on hold
• Trace forward from any combination of manufacturing factors to identify products made under those conditions and impacted customers

You can make better manufacturing decisions when you leverage Machine Learning and Artificial Intelligence capabilities. Building this from scratch is possible, but it will take a while. Instead, it is better to leverage a ready-to-go Adaptive Intelligence Application that was specifically made for Manufacturing. This will help turn data into actionable information quickly, as it lets you analyze key patterns and correlations that are related to manufacturing operations, predict the probability of critical events and take proactive measures to address them, trace manpower, machine, material, method, and management related information, and identify impacted products, processes, suppliers, and customers and take actions to mitigate risks.

To find out more about Oracle Adaptive Intelligent Apps for Manufacturing, click HERE.