Optimising manufacturing productivity through Condition Monitoring

When it comes to manufacturing assets or production lines, it’s important that operations be as efficient and productive as possible. Stephen Hayes, MD of Beckhoff Automation, discusses the benefits of condition monitoring.

At first glance, Condition Monitoring (CM) sounds relatively simple. It is the process of monitoring a set of parameters in a physical asset, to ascertain if that asset is performing within the specified tolerance. Through CM, we can analyse key parameters to help improve efficiency and/or output, and identify or even predict developing failure modes. In other words, routine monitoring of your manufacturing and production equipment provides you with vital data on factors like vibration, temperature and noise that provide an overview of the current condition of your equipment. CM is a core component of ‘predictive maintenance’, helping companies to detect developing faults, reduce repair costs and minimize the need for scheduled and unscheduled downtime. Moreover, effective asset monitoring positively impacts asset safety, reliability, longevity and, ultimately, productivity.

Condition Monitoring is a prime example of where real-time data can be used to drive productivity. Of course, this generates very sizeable and complex data volumes. For asset analysis to be accurate, the data it’s based on must also be accurate. Without quality, reliable data, companies face the costly scenario of “garbage in, garbage out”. The choices regarding whether to consolidate and analyse that information locally (through embedded hardware) or stream it to a data centre can also have a significant impact. The relatively low cost and accessibility of computing power – both in-house and externally – has meant that digital programmes and practices are now core features within the industry. This is especially true for high-value engineering and manufacturing enterprises….. And yet there’s a level of resistance to digitalisation.

One barrier is scale. A company-wide strategy often needs to be implemented, and the change in practice may result in the need to retrain engineers in the field. However, I tend to compare this to the days when typewriters were being replaced by PCs. The companies that were the last to give up their typewriters were the ones losing out on opportunities – they simply lacked the capabilities of their competitors with sophisticated IT departments. It’s worth stating that no one in the industry is trying to force change, but the reality is that companies who don’t implement digitalisation programmes won’t be able to compete. And without an integrated digitalisation program, CM won’t be as effective.

Our experience at Beckhoff Automation dates back to 1980 when German physicist Hans Beckhoff had the idea that personal computers could function as industrial control systems. For that to work, the validity of the data would need to be preserved between the field and PC, and the PC itself would have to operate in a deterministic, reliable way. Moreover, this activity would need to take place in an environment that industrial software engineers and controls engineers could program with ease. These criteria were met, and the result was the first PC-based machine controller. By integrating the CM application into a PC-based controller, you have one physical system that controls the machine and allows you to run monitoring functions alongside control functions, without incurring the cost of additional hardware or software. Our proprietary software can take advantage of CPU performance by using all the different cores of a multicore processor, and dedicating a core to condition monitoring.

We’ve been working with the Centre for Modelling & Simulation (CFMS) on integrating our systems into new techniques and practices for production. As an independent, not-for-profit specialist in high value design and manufacturing capability, CFMS has a number of systems that generate large volumes of data. By demonstrating capture and analysis of this data using a digital framework, CFMS has been able to evaluate the cost benefits of these technologies before they’re deployed in a live production manufacturing environment. We created a technology demonstrator with CFMS, involving linear motor single-axis robot. The variables were connected directly to the CFMS cloud infrastructure and our data scientists ran artificial intelligence programs that simulated rules of energy tariffs during the day for a factory. We factored in actual production requirements, and were able to provide the optimum way of controlling the machines based on those simulated rules. Production output was maintained, and the machines were operated during the cheapest tariffs.

By working with CFMS, we’ve been able to explore how artificial intelligence based on outside rules can affect machine operations, from costs to longevity. We’re now looking at running predictions of component failure, and feeding that back into the system. Anomaly detection is another scenario where algorithms are fed with historical data enabling engineers to see when a machine is about to deviate from tolerances. Using CFMS’ resources and expertise, there are many ways we’re exploring the true potential of CM. Beckhoff implements open automation systems based on PC Control technology, for more information about the themes discussed, contact Beckhoff.

If you’re interested in exploring how an integrated digitalisation strategy can transform your digital engineering capabilities, contact CFMS for further information.

Sign up to our newsletter