According to a survey conducted by BearingPoint, predictive maintenance in industry has been discussed more than it has been implemented. While 84 percent of respondents from the mechanical engineering, chemical /pharmaceutical, and automotive industries deal with the subject of predictive maintenance in their company, only one in four companies has actually carried out initial projects. There is a lack of courage to make mistakes and learn from them, says Donald Wachs, Global Director of Industrial Manufacturing and Industry 4.0 at BearingPoint.
What are the biggest obstacles to using predictive maintenance?
In addition to IT security, according to our study the high implementation costs and availability of data are the main reasons why predictive maintenance is only gradually gaining ground. In particular, there is often a lack of truly relevant data from intelligent products and assets to assess the prediction of conditions. The complexity of the solutions also plays a role. There are hardly any finished solutions. Therefore, companies have to develop individual solutions with appropriate partners. And last but not least, management often fails to commit.
When does predictive maintenance pay off?
For the most part, use cases are only designed for small projects, but at the same time one hopes for early amortization. However, a predictive maintenance project often pays off in the context of major roll-outs or a large number of integrated intelligent systems. In addition, predictive maintenance cases should be enhanced with additional digital products to support economic efficiency.
What do you advise companies: How should they approach predictive maintenance projects?
They should develop a real use case to avoid the risk of creating data pools without added value. Depending on the goal, the starting point should be a well-defined and relevant use case. For example, the desire to increase equipment availability or reduce costs. This involves virtual inspections or remote maintenance. If the first tests are successful, they can be expanded successively. Companies may also wish to improve customer satisfaction. For this purpose, they can provide customers with status information or incident histories for the machines they use via a Web portal. This means: Start with a pilot to learn and test. Consider aspects such as data connection, data analysis, and holistic and economic integration into maintenance management. Last but not least: Learn from mistakes instead of planning every detail in advance.
Excessive complexity is another reason for the cautious use of predictive approaches. Are the solutions really that complex?
Every solution has to be tailored to the individual needs of the company. Although there are blueprints available for certain components, the behavior and wear within a specific system are always different. In particular, the algorithms require special consideration and continuous improvement. The validation of their respective parameters, considering individual impact factors for detecting anomalies, still presents a methodological challenge in the context of a superior prognosis.
Every year around four million cell phones disappear in Germany. With every theft or loss, the owners face expensive replacement costs. A blockchain for smartphone identification numbers could facilitate the locking of mobile devices.