Predictive analysis and machine learning: The future of manufacturing maintenance
By Dave Hewlett Hitachi Solutions
By Dave Hewlett Hitachi Solutions
Mar. 31, 2017 – The Internet of Things (IoT) is having a major impact on businesses across almost every sector, improving efficiencies and increasing output. The manufacturing industry is no exception; in fact, it may be one of the most affected.
Per a Tata Consultancy Survey, manufacturers utilizing IoT solutions in 2014 saw an average 28.5 per cent increase in revenue between 2013 and 2014. It comes as no surprise then that manufacturers are investing heavily in IoT technology.
By investing in advanced technology, manufacturing corporations can take on the most pressing challenges facing the industry today. For example, increasing capacity and efficiency makes it necessary for companies to reduce equipment downtime, lower maintenance costs, and address aging equipment. Similarly, there is increasing concern surrounding worker safety and environment. IoT and predictive maintenance will play a key role in addressing these concerns and will be essential for manufacturing businesses that wish to compete and grow in the current market.
This article will discuss how the IoT is enabling machine learning and sophisticated predictive analysis that addresses the latest manufacturing concerns — and gives businesses the ability to transform their equipment service management operating models. But first, let’s briefly define the technologies that are transforming manufacturing.
• IoT: The IoT is a network of connected devices that are able to collect and exchange data using embedded sensors. The IoT, while still in its infancy, is the result of advances in Cloud technology, big data and ubiquitous data networks.
• Machine learning: Machine learning is an application of artificial intelligence (AI) that gives a machine the ability to access data and “learn” or act for itself without being explicitly programmed.
• Predictive analytics: Predictive analysis uses data, statistics, machine learning, AI and modelling to make predictions about future outcomes. When it comes to manufacturing, predictive analytics offers the ability for businesses to transform from a repair-and-replace to a predict-and-fix maintenance model — this is known as predictive maintenance.
Predictive maintenance and machine learning are changing manufacturing
An industrial machine breakdown can delay processes and cost manufacturing business much more than the cost of repair. Predictive maintenance technologies allow manufacturers to move from a reactive maintenance model to a proactive maintenance model, giving them critical information about when and where outages are likely to occur so they can prevent downtime and increase asset utilization.
This new technology, supported by the IoT is creating a more efficient environment by both increasing productivity and decreasing downtime. Some predictive service success stories report reductions of downtime by as much as 50 per cent, while reducing maintenance costs between 10 to 40 per cent. This significant reduction in costs boosts profitability for manufacturers while reducing capital investments into new equipment by 3 to 5 per cent.
“In manufacturing, these savings have a potential economic impact of nearly $630 billion per year in 2025,” according to a McKinsey & Company report.
How predictive analysis works
Predictive analytics are powered by the data collected from the devices or sensors embedded in a manufacturer’s industrial equipment. This equipment, with their embedded sensors, then becomes part of the IoT, actively collecting data which it shares via the Cloud, enabling machine learning and predictive analytics. Using this raw data — that is collected, stored and analyzed over time — manufacturers can identify insights and trends that can improve business performance.
This collected data is used to mobilize predictive maintenance capabilities that can anticipate production disruption and monitor processes and assets remotely. Asset management can be greatly enhanced by syncing up production schedules and parts availability to ensure the greatest level of efficiency. And corrective actions can be issued before an industrial asset breaks down, allowing service to be performed preemptively and the results recorded, so that further data analysis can be conducted to spot trends in asset life and maintenance. Finally, with a portal that offers inclusive cross-organizational dashboards, teams can collaborate with ease and make informed decisions, improving output and efficiency.
Benefits of predictive maintenance for manufacturing companies
Increase equipment uptime
With predictive analytics, manufacturers can reduce equipment downtime by proactively monitoring equipment health in real time so service can be performed exactly at the time needed, ideally before it fails. Similarly, productive uptime can be increased by predictively identifying problems so repairs can be performed during scheduled production downtime rather than during peak periods. By monitoring equipment in real time, manufacturers can avoid costly breakdowns that can interrupt production and ensure equipment is always operating at optimum performance.
Improve quality of service delivery
The IoT and predictive maintenance also offer the opportunity for manufacturing companies to cut routine maintenance costs. By performing condition-based maintenance that addresses actual issues rather than performing costlier rule-based maintenance that misses developments between intervals, manufacturers will realize significant savings. Similarly, by performing predictive maintenance before catastrophic failures occur, asset lifespan can be increased, keeping aging equipment operational.
Create new equipment service revenue streams
Utilizing the power of predictive maintenance, equipment manufacturers can create new revenue streams by enabling performance based service agreements or equipment-as-a-service offerings. For equipment manufacturers or resellers, this means they can transform their business model from selling equipment to equipment-as-a-service with performance guarantees, and new contract-based predictive maintenance service offerings can be offered by equipment manufacturers and third-party service organizations. As well, sensor data gathered by equipment manufacturers or third-party service companies can be used to identify cross-sell or up-sell opportunities.
Improve worker safety
Worker safety is a major concern for many manufacturing companies. By leveraging machine learning and predictive analytics, manufacturers can improve worker safety through the monitoring of equipment conditions and faults. When an issue arises, the machine can produce an alert or issue counter measures before a human injury is sustained.
Manufacturing is certainly on track to see major changes in the coming years, as more and more manufacturing businesses invest in predictive analytics and machine learning technology in an effort to transform their operating models.
Dave Hewlett currently serves as vice-president, Business Intelligence and Advanced Analytics, Hitachi Solutions. To learn more about Hitachi’s IoT and predictive maintenance solutions, please click here.