Predictive maintenance is the new big thing. While big data provides the ability to collect a vast amount of information, nobody really expected it to radically change the way we produce things. But using big data to predict when, where, and even how a tool may malfunction can make a major impact to a business’s bottom line.
Unplanned machine failures are estimated to cost manufacturers in Great Britain an excess amount of £180bn every year. For individual businesses, this can vary from a few thousand pounds in the fast-moving consumer goods industry up to millions of pounds in the automotive sector.
Big data has always held the promise to bring efficiencies to the way we manufacture and produce goods and bring new insights to the whole production process. But the reality is, it can go much further than that and can be used to totally re-write business models. Big data is now enabling a paradigm shift in the approach to manufacturing in a whole range of industries.
One of the main benefits big data provides is forecasting when a machine or tool needs attention before it stops working as designed. This is predictive maintenance. It means an operator can plan maintenance during a period of scheduled downtime, before a real problem occurs − lowering maintenance costs as well as providing an increase in overall equipment effectiveness.
Cost savings with predictive maintenance in production
Using the power of historical and real-time big data to feed predictive analytics results in forecasting models that predict when a machine will fail, or when it will fail to produce the optimum quality of output. Calculating a tool’s useful life allows companies to plan their maintenance budgets and their production schedule in advance. If a machine is not working, it is not making money for the business. Being able to predict and prevent machine failure is therefore a huge advantage to businesses.
At the same time, companies can factor in the savings from any potential damage to a manufactured product.
By using predictive maintenance, companies can save money in two important ways:
1. Focus on the predicted maintenance times rather than fixed, stipulated service intervals. At Industrie Reply, we estimate that this can reduce maintenance costs by between 10-40 percent. Additionally, planning improves efficiency and we estimate the maintenance team is 45-55 percent more productive. At the same time, the precision of the maintenance procedure improves by up to 85 percent.
2. Predict down time as well as prevent unplanned machine outages, allowing maintenance work to be conducted ahead of time. With a reduction in unplanned outages, overall equipment effectiveness is increased by between 30 to 50 percent.
Although cost savings are always appealing, in the past it has been costly to achieve those results once you factor in the necessary investment in digital analytics tools. But over the last few years, operating costs have drastically reduced. Most production machines today are already equipped with the sensors needed for predictive maintenance tasks. It means there is no longer a need for the cost-intensive retrofitting of sensors. On top of that, the costs of data storage and data processing has reduced over the years. Businesses now have access to easy-to-use forecasting models, so they no longer have to rely on highly trained service technicians with specialist statistical knowledge.
Development costs have also dropped significantly in the last few years. Consultants now make use of pre-trained machine learning models, which are based on many years’ experience through the use of similar or identical machines, such as drives and pumps, and components.
What’s more, businesses can now deploy predictive maintenance models much faster than previously, so they can achieve a faster ROI (Return on Investment).
New service models with predictive maintenance
Predictive maintenance offers a range of additional benefits. By offering this as part of their digital services portfolio, mechanical engineers can cut down on machine outages, whilst optimising the maintenance work required.
In addition, stock levels are highly reduced as parts are only ordered as and when they are needed. This frees up space for other, potentially revenue generating items.
The improvement in service provision has a range of other tangible benefits; customer loyalty improves as interaction increases. In the durable capital goods industry, the touch-points between mechanical engineers and customers is even expected to increase hugely.
Engineers can also provide a better quality of service as resource planning is enhanced and maintenance dates are agreed upon well ahead of time. Finally, the product improvement process is likely to accelerate as an understanding of the problems and defects improves.
It’s a quadruple whammy of improvements; more frequent interactions; timely service visits; higher quality; and in a shorter timeframe.
Predictive maintenance is a pioneer for new service models such as “Power by the Hour”. This means, for example, predicting the provision of forklifts per hour, forming by the hour, pressing per hour, and so on. Customers only pay for the services they need and use and can consequently avoid high investment costs. Machinery suppliers experience fewer unplanned machine outage times and minimise hits to their income streams.
Furthermore, after-sales teams can provide more services to their clients by adding predictive maintenance-as-a-service for a machine to their portfolio. Meanwhile, forecasting models with intuitive graphic user interfaces are readily available, reducing the overall need for specialised service technicians.
All in all, a company with predictive maintenance can be far more proactive and react to not only its own needs, but to those of its clients too. Maintenance will be less costly and result in less downtime and the relationship with clients improves as direct contact increases and services are more tailor-made and timelier.
Big data is a new frontier and with new capabilities come new opportunities. The future looks bright as we usher in a new era of productivity for manufacturers and new, stronger relationships between manufacturers, their suppliers and their customers.
Clemens Weis, Partner and Christoph Schmierer, Manager at Reply
Image Credit: MNBB Studio / Shutterstock