Smart factories and the rise of predictive maintenance
December 3, 2018 – Smart factories are rapidly becoming the future of manufacturing, offering a new level of efficiency and productivity to those investing in them.
Industry 4.0, combined with increasingly sophisticated analytics, is playing a huge role in driving the smart factory movement.
Manufacturing executives and engineers no longer see a factory as a mass of machinery operating as part of one or more individual production lines. Instead, they see an interconnected network of moving parts, something more akin to a living and breathing organism, that can be fine-tuned to optimize performance.
New technologies such as big data, prognostics, artificial intelligence (AI) and the cloud ensure that those who manage and maintain manufacturing environments can get on the front foot and proactively manage these environments, with relatively low levels of investment.
The potential of this new era of responsive machinery is expansive, enabling a higher level of communication, transparency and, therefore, yield between all stakeholders. Business owners can predict and manage the failures of their machinery and, with a new call for servitization, suppliers can minimize their maintenance costs.
The rise of predictive maintenance
For decades, condition monitoring and predictive maintenance have been a reality for the few and a dream for the majority. The cost and complexity of gathering and analyzing sufficient data to drive tangible results has limited predictive maintenance to the defence and aerospace sectors. However, the rise of the Industry 4.0, where production assets are connected to the internet and able to communicate, combined with AI and advanced analytics, has opened up the world of predictive maintenance to a much wider range of industries and use cases.
Until recently, gathering the data required to inform conditioning monitoring activities has been a laborious manual process requiring specialist expertise. The growing use of self-sensing machines that record their own vital statistics and relay them for analysis over the internet, and the ability to retrofit older assets with smart sensors to do much the same, allows data to be pulled from potentially thousands of machines at relatively low cost.
The concept of a smart factory is entirely dependent on the connectivity enabled by Industry 4.0. Machines that can sense and communicate can provide a vast amount of valuable data. However, this data needs to be filtered and analyzed if it is to be translated into actionable insight for manufacturers.
Much like the gathering of machine data, historically this analysis would have been a highly manual endeavour, involving teams of expensive data scientists. More recently, however, organizations have developed intelligent software to automate this activity, creating a bespoke algorithm to identify problems and, crucially, spot the signs that indicate if and when a machine will fail in the future. This prognostic approach allows engineers to undertake precisely the right maintenance activities during periods of planned downtime and fix problems before they can affect production.
Bespoke algorithms are important because the condition data outputs from machines – even two of the same make and model – are as unique as a human fingerprint. AI does the heavy lifting here, fine-tuning the performance of each algorithm to maximize its accuracy.
Industry 4.0 is a vital part of this process. Automation is only possible because of the data that is gathered from the machines, but it also requires computing power to analyze that data. This crucial analysis is performed elsewhere on the internet, in the cloud, where the resources exist to power the AI and run the algorithms on a continual basis, regardless of where the machines are located.
The benefits of reducing downtime
Recent developments in condition monitoring and predictive maintenance, made possible, in part, through Industry 4.0, are significantly improving the performance of industrial machinery, while making the task of operating and maintaining it easier and more efficient.
The cost of unplanned downtime is a huge drain for any manufacturing environment. In the automotive sector, for example, each minute that critical machinery is offline will cost the factory tens of thousands of pounds.
Gathering data and connecting with services in the cloud are already driving significant improvements in machine effectiveness and efficiency. The longer-term potential of these technologies has yet to be realized, but it is clear that it will be truly transformative.
Dr. Simon Kampa is CEO and co-founder of Senseye, a predictive maintenaince solutions provider.
Level II Introduction to Robotics – e-Cobra/Viper/Cobot
July 9-17, 2019
Product Safety & Liability Prevention Seminar
August 7-8, 2019
Digital Industry USA
September 10-12, 2019
EMO Hannover 2019
September 16-21, 2019
Autonomous Mobile Robot Conference
September 17, 2019
International Metrology Congress
September 24-26, 2019