How data, Industry 4.0, and AI are shaping modern asset management

Published:  16 November, 2021

How data is stored, collected, managed, and analysed, dictates the success of asset management from company to company. From condition monitoring to informing a spares strategy and even predicting which assets are needed in the future, data, when used well, can provide indispensable information for organisations of any industrial sector. Thomas Boswell* examines how digital industrialisation is transforming MRO and underpinning proactive asset management strategies.

Condition monitoring Condition monitoring essentially informs an asset management strategy by improving a business’ understanding of its assets, how they are functioning, and whether they are detrimental to productivity in the long, medium or short term.

It is more than just sticking a sensor on a machine. It’s a consistent evaluation of machine performance over time. In the same way that when a patient’s heart rate, temperature or other readings deviate from the norm, data associated with failure modes of machinery can indicate the health of assets. This data, such as variations in vibration, temperature, or noise, can be collected manually or automatically using sensors and telemetry solutions. Its analysis provides valuable insight into what is, or is not, working as it should, to inform what remedial action should be taken.

For example, we had installed and commission a system to investigate a bearing failure on a 5000 tonne multistage press, which had resulted in two days’ worth of downtime at an automotive manufacturer, resulting in a loss of around £80M in lost production. The press is the largest of its type in Europe and stamps sheet aluminium into vehicle panels. The condition monitoring system monitored vibration data associated with the most prominent failure modes of the press’ main drive motors, gearboxes, and flywheel / clutch bearings, based on predefined setup properties and alarms. A diagnostic unit was also integrated, and the result from the algorithms within the units were transmitted via an Eriks eConnect product to a cloud hosted trending and alarms solution where the customer had a unique log-in and data is available 24-7.

The system began to alarm continuously on the second fly wheel bearing after levels of vibration breached the yellow warning, eventually breaching the red critical alarm a month later. The data showed a defect on a fly wheel bearing which had a 630mm bore size and costs £25k for the thrust assembly. Once the bearing assembly was replaced, the levels of vibration returned to acceptable levels.

Providing foresight

Equipment used on the factory floor will require spare or back-up components, to minimise downtime and make sure that any faults or breakdowns are dealt with efficiently and quickly.

Without an informed spares strategy, businesses are left with one of two options: either they stock everything they could possibly need, or they make reactive purchasing decisions based on their needs of the moment. Neither of these are ideal. The first, because it places constraints on budgets, space and staff. The second, because it relies on stock being readily available from third-party suppliers, without factoring issues such as obsolescence or delivery.

Historically, a spares strategy is normally dictated by the criticality of assets and past failures. However, the criticality of an asset becomes dynamic when you factor in the condition. The level of insight provided by condition monitoring data means that businesses can identify critical components that are most likely to fail. The777se, along with key supplies that are difficult to source, can be taken into account when deciding which spare or new parts should be prioritised for purchase. At the automotive manufacturer for example, the condition monitoring helped to identify the fly wheel bearing as a critical component. As a result, the customer purchased two more spare fly wheel thrust assemblies at a significant expense, to be put in storage.

Once in a storeroom, a database can collate all the spares available, monitoring life expectancy, condition and rotation to ensure that spares are maintained in a productionready condition.

Challenges ahead

Whilst a reactive MRO approach may suffice for a non-critical asset which can be easily and cheaply fixed or replaced, the more complex and critical an asset is, the more data and advanced skills are needed to monitor its condition, predict required maintenance, and diagnose problems accurately.

The ability then to process and analyse these numerous streams of information simultaneously – even in real-time - will grow in importance. This is where AI will deliver the greatest benefits. Its capability in analysing information as well as learning and adjusting to new inputs will inform business decisions to improve efficiencies and reduce wastage across the supply-chain.

However, there are still formidable barriers to implementing Industry 4.0 and AI in a way which will really benefit MRO. Firstly, businesses need to attract those with the right skillset and upskill their existing workforce so that employees can operate and work with digital systems at the level required for their roles.

Secondly, the challenge is well-known in terms of interoperability. Industry must develop standards for which technology is bound, enabling the easy transfer of data across platforms. This is one of the biggest challenges faced by AI, as without a common language, analysis of data across platforms will be incredibly complicated as the information sources and languages are potentially endless.

The success or failure of AI is also underpinned by the seamless transfer of data between equipment and suppliers, and so ensuring ease of access to data will be key. But many OEMs, component suppliers and manufacturers remain deeply reluctant to do this over security concerns. ERIKS’ own research shows that 79% of respondents would only share very limited information with their OEM equipment partner, if at all. In the minds of IT directors, the risks of opening up IT networks to suppliers outweigh the benefits.

Industry therefore needs to start breaking down the barriers that are going to prevent data sharing. Developing appropriate security solutions, from firewalls to private cloud environments, FOG computing and beyond will address these concerns, but we must also promote the benefits that greater data sharing can bring.

Conclusion

Data collected by condition monitoring is helping us to identify patterns, changes and warning signs before they become a problem. Not only can organisations stay one step ahead of their assets at all times, but also reduce the likelihood of downtime, damage, or loss of production. The level of insight provided by data delivers benefits beyond predictive maintenance too. It is informing organisations’ indirect supply chain and procurement practices. The impact this could have on a business should not be underestimated. Afterall, this kind of proactive foresight will define asset management strategies of the future. Data, digital technology, and Industry 4.0 is making all this possible.

For more information, please visit: www.eriks.co.uk

* Thomas Boswell is Eriks’ reliability engineering manager

https://twitter.com/ERIKS_UK

https://www.linkedin.com/company/eriks-uk/

Sign up for the PWE newsletter

Latest issue

To view a digital copy of the latest issue of Plant & Works Engineering, click here.

View the past issue archive here.

To subscribe to the journal please click here.

Poll

"How is your manufacturing business preparing for a net Zero target?"






Twitter