Podcast: Six tips to leverage machine learning in manufacturing

The importance of machine learning

Hollywood portrays artificial intelligence (AI) and machine learning (ML) through far out sci-fi and action movies; however, applications are becoming much more common in everyday life. From machine learning algorithms recommending which movie viewers might enjoy based on prior habits to chatbots efficiently assisting customers to the content, salesperson, or service page they need, and sports franchises using machine learning to help sign and trade players, the future of machine learning appears to be here. For manufacturers, the real-life business applications of integrating ML/AI data to make more informed decisions can be vital explains Paul Dunn, VP of product management at Infor, in a recent podcast interview with Panorama Consulting Group.

Manufacturers of all sizes can gain value from these advanced (yet accessible) modern technologies that leverage data insights and easily integrate into existing systems. Dunn sees AI/ML for manufacturers as particularly valuable in the current environment where supply chain volatility and anomalies continues to challenge the industry. With all the uncertainty, manufacturers are turning to machine learning solutions, especially for predictive quality control and product maintenance to ensure they are producing good products, monitoring system maintenance to avoid production disruptions, and capturing their internal knowledge.

Dunn’s guidance for manufacturers considering how to implement AI/ML: “Simple is better. You do not need to be complex; however, the intelligent models do need to be able to adapt variables to deliver real value.” A recent example that highlights the value of ML/AI is with the COVID pandemic that threw a huge curveball at manufacturers who had to quickly figure out how to adapt. Demand forecasting with the benefit of ML allowed manufacturers with those tools to quicky, and more accurately, determine what needed adjusting.

“It is important to start with simple processes and analyses, and then branch out based on those learnings with more advanced AI/ML. It may seem counterintuitive, but start simple and build upward, rather than start complex and build downward,” shares Dunn who explains that most enterprises already have significant amounts of data within their ERP, CRM and HR maintenance management systems and manufacturing execution systems.

In the podcast, Dunn discusses six primary value drivers for manufacturers who integrate AI/ML:

1. Process Intelligence – Improves efficiency and decision making for both business and manufacturing processes. This can be as simple as how do I detect anomalies in our accounting system to prevent fraud? Information like this is simple for AI/ML to detect.
2. Asset Intelligence – Very commonly used for predictive and preventative maintenance. It is helpful to be able to predict when a piece of equipment might not work as it is supposed to or have unexpected downtime which is one of the leading causes of lost opportunity and lost value for a company during production.
3. Forecasting Intelligence – A newer tool that is necessary for companies to help with supply chain issues. Helps with knowing what/when to order materials and can even predict when some materials may be short.
4. Sales Intelligence – Helps with improving efficiency and effectiveness of customer relationships, answers questions like: Who are my best potential target customers? As well as helps with forecasting sales goals for the coming year.
5. Pricing Intelligence – Supports market-centric pricing and detects pricing anomalies or opportunities in the market where prices could be raised to increase margins as well as where margins may be shrinking and how to deal with that.
6. Human Capital Intelligence – Provides labor efficiency with insights about how to improve employee satisfaction and retention. Can identify employees who might be at risk before they quit. This is important with today’s tight labor market.

It is important for manufacturers to start thinking about what kind of information they need to deliver a certain result and then build in the AI/ML tools with semantic data models to generate accurate and helpful results.

Dunn’s concluding thoughts: Do not go it alone when you are determining how to utilize and integrate machine learning or artificial intelligence for your organization. It requires focus and skill sets from those with experience in these areas. AI/ML can be valuable assets to expand a manufacturer’s knowledge about its processes, products, and production. Integrating AI/ML strengthens strategic decision making and can deliver bottom line value.

To listen to the podcast interview, click here.