These are challenging times in manufacturing. With an election imminent in the U.S., uncertainty in the Eurozone, and a fragile economic recovery, it’s a risky time to be investing in technology. But with intense global competition, who can afford to delay? Manufacturing is in the middle of a technological renaissance and standing still could equate to business suicide.
Despite these challenging conditions, research by Deloitte and MHI indicates that one particular technology – predictive analytics – is set to undergo rapid adoption as the most important advance that manufacturers need for future competitiveness.
Predicting the future of manufacturing
In recent years the Internet of Things (IoT) has revolutionized manufacturing. The factory floor is now networked and is being linked into the supply chain, giving managers visibility into the performance of their manufacturing equipment, inventory levels, and the intake and depletion of raw materials.
But clever use of all this IoT data can achieve even more. It can be used to predict the future: what customers will buy, which equipment will fail and which suppliers will miss delivery deadlines. With such predictions, manufacturers can intervene and change the future before it actually happens – reducing production costs, improving customer service, and leap-frogging the competition. Predictive analytics is the potent ingredient needed to turn manufacturing data into these valuable insights.
Algorithms: the secret sauce
Predictive analytics software anticipates future outcomes by using machine learning algorithms that find patterns in large volumes of data, which in turn enable predictions to be made. In a recent Digitalist blog, How a Little Bit of Wizardry Can Transform Your Bottom Line, I explained how these predictive models are being be used to automate decision-making – sometimes between two or more business processes. Predictive analytics is transforming manufacturing in a number of areas:
Supply chain optimization
With unrelenting pressure on manufacturers to improve speed to market, meet the demand for customization, and work to “lot sizes of one,” optimizing the supply chain is more relevant than ever. Predictive analytics helps by combining and analyzing factory information with supply chain data like raw material availability, customer demand, and the delivery of finished goods. The results enable better inventory management, more reliable transportation networks, and less variability in lead times. These benefits are so significant that 44% of respondents to the MHIs survey said that predictive analytics have the potential to be a disruptive or competitive force in their industry.
As an example, a global manufacturer of construction and mining equipment has mapped more than 10,000 suppliers to risk scores to determine which suppliers offer value and which ones pose risk. As a result, they have been able to reduce unnecessary costs from supplier disruptions and have improved management of inventory levels. Elsewhere, a chemicals business has boosted its bottom line by using predictive analytics to enhance their demand-based forecasting – enabling the company to meet customer needs quickly without overstocking its inventory.
Early fault detection
Today’s consumers expect product personalization and continuous innovation, causing manufacturing headaches due to growing product complexity and the increased potential for faulty products. In industries with long production lead times or potentially harmful or unsafe products, early fault detection is critical. The automotive industry is a prime example where the cost of recalls can be devastating. In the UK alone, 1.45 million vehicles were recalled in 2015 as per DVSA 2015, and in the United States during the same period more than 51.2 million vehicles, an all-time record.
In my earlier blog, I highlighted a leading automotive business that is using algorithms to process 30,000 data sets per second from up to 100 engine components to predict engine failure as early as possible – minimizing lost hours and resources. Automotive manufacturers are also using predictive analytics to combine internal and external data to understand what is happening in the after-market; spotting which parts are failing and predicting the likelihood of a recall. By doing this, automakers can be quicker and more proactive in responding to product quality and safety issues, resulting in better recall management.
Thanks to embedded sensors channelling operational data from machinery to the IoT, organizations are now constantly monitoring the health of their manufacturing equipment and using algorithms to predict the likelihood of failure. This allows preemptive action to avert imminent failure and reduces unplanned down time. The result: better return on manufacturing assets.
Organizations can also move away from reactive or scheduled service models and service machines only when needed, resulting in significant cost savings. In Germany, a compressed air systems and services company monitors over a million measurements a day at each of its clients’ compressors. It then uses predictive analytics to predict the health of these machines, and in doing so enables its clients to avoid unplanned downtime, decrease time to resolution, and increase optimal equipment longevity and performance.
For manufacturing businesses that have invested in the IoT, predictive analytics is the important next step towards minimizing production costs, optimizing supply chain, and maximizing factory uptime. Some manufacturers are taking it even further – using it as a source of innovative new services and revenue streams. As a route to gaining a competitive edge, predictive analytics is an investment that should be made today, not tomorrow.
For 6 important points about algorithms, read Algorithms: The New Means of Production.
To learn how your business can embrace predictive analytics, read Predictive Monthly: The Role of Predictive in Your Analytics Strategy.
Want to unlock the real value in your data? Click here to learn how analytics can help you make smarter decisions.