AI, or artificial intelligence, and advanced data analytics have dominated the manufacturing operational technology (OT) discussion for the last several years. Whether talking about maintenance, quality, or manufacturing operations the media and supplier community have been extolling the benefits of using AI-driven analytics to shift — from reactive behavior to performance based on better insight-guided predictive behavior. Larger manufacturers started making the shift to predictive maintenance and then predictive quality about five years ago. Small and mid-sized businesses joined that trend as they adapted to the market upsets during the COVID pandemic. Today, thanks to the broad use of this technology by most suppliers, predictive maintenance and quality tools are available to virtually every manufacturer, regardless of industry, size, or location. The next major shift coming to the industry is two-fold. Operational analytics is now becoming mainstream, and more importantly, AI is enabling the next generation of analytics which will elevate decision-making from a predictive nature to one that is prescriptive.
The Analytics Continuum
Ever since the digitalization of manufacturing started 50 years ago, the increasing availability of information about operational performance has evolved. With the advent of data historians, analyzing past performance to better understand and diagnose both operational failures and exceptional performance has been the initial step all manufacturers take when using the data they had access to. Whether trying to replicate a “golden” batch or production run, or to define alarms or predict an impending failure or upset, the goal was to understand the past and predict possible future outcomes. The challenges with this approach have always been that it requires a historical event to define the conditions that you are seeking to detect, in many cases, the relationships being relied upon, are based on correlation, not causation. Consider the example of when you have a front tire fail as you are driving. Before there were tire pressure monitoring systems (TPMS) in most vehicles the first sign that you might have an impending failure was that the handling of the car became somewhat erratic. Generally, the failure was caused by a loss of air pressure in the tire leading to excessive flexing and overheating which ultimately causes the blowout. With TPMS now a standard feature on most cars, the warning of low air pressure alerts drivers well before they have a serious handling issue that there is a problem and for many of us, tire failures are a far rarer occurrence than they were for a previous generation of drivers.
The ability to predict future behavior has transformed the quality and reliability aspects of manufacturing and is poised to do the same for other aspects of operations such as production speed and volume. A modern process planning and execution system, also known as a Manufacturing Execution System (MES), with advanced analytics, is capable of providing guidance to operators and supervisors on how to optimize production throughput while maintaining quality and achieving sustainability goals. Today’s Manufacturing Execution System isn’t just a reporting and control application but is a powerful tool for manufacturers to achieve and sustain competitiveness. With a richly featured MES, with AI and machine learning capabilities, manufacturers, regardless of size, can optimize performance better than at any time in the past.
Moving from Predictive to Prescriptive
Many large manufacturers are already taking the next step in their adoption of reliability analytics by going beyond predictive maintenance to prescriptive maintenance. Modern reliability tools with AI/ML-based analytics cannot just warn of an impending failure but suggest operational changes to either delay or prevent that failure. For example, by monitoring bearing conditions such as temperature and vibration, rather than just providing a warning that a bearing is heading towards a failure and maintenance is soon needed, modern prescriptive tools might suggest that by slowing down by 10% bearing life may be extended to the next scheduled maintenance or replacement interval.
The future of MES analytics is headed the same way. Thanks to the ever-increasing amount of data available from production equipment due to the growth of the industrial internet of things (IIoT) and wireless data collection, the modern MES is becoming capable of providing the same type of prescriptive guidance to other aspects of quality, operations and shop floor performance. Large manufacturers have had the luxury of being able to afford large analytics groups to do these analytics and optimize manufacturing performance at a corporate level. With the growth of prescriptive capabilities in the modern MES, even small and mid-sized manufacturers now have access to tools that allow them to see the same types of performance gains as larger enterprises.
As a Senior Director of Product Management at iBASEt, Attila brings over 12 years of Product Management and 10 years of Software Development experience to the leadership team. He is responsible for overseeing the product management operations across the organization and ensuring that the procedures meet the business requirements and client specifications. He also has deep experience in developing product strategy, roadmap, and requirements with cross-functional teams and in launching successful software products.
Attila earned his BS degree in Computer Sciences in Hungary, holds a IPMA Project Manager certification from Germany, and has contributed to over 15 patents.