What happens next? Maintenance must transform if manufacturers are to meet their Industry 4.0 goals

New opportunities are emerging for manufacturers to embrace Industrial IoT and to transition maintenance operations from a reactive to a predictive model.

Businesses need to act now to evaluate how new technologies fit into their long-term manufacturing strategies and capitalise on these developments – or risk getting stuck on the wrong side of disruption. Kevin Bull, Product Strategy Director, Columbus UK, explains how predictive maintenance is transforming manufacturing and which technologies businesses need to focus on to stay ahead of the competition.

There is a huge push towards Industry 4.0 with manufacturers of all sizes looking to embrace emerging technologies such as AI, machine learning and advanced data analytics as part of their digital transformation. In our recently published ‘Manufacturing 2020’ report, companies cited production flexibility, reduced costs and increased output as the long-term benefits of improved factory connectivity. The evolution of asset maintenance is going to play a big role in achieving these goals.

Developing an IoT vision of manufacturing

Industrial IoT (IIoT) provides the framework on which further disruptive technologies can build and contribute to improved maintenance practices. IIoT deployments monitoring temperatures, vibrations or humidity from sensors embedded within equipment on the plant floor, all generate large volumes of data in real-time. This is uncovering new insights into processes and sub-processes we’ve never been able to capture before and dramatically changing how we schedule and predict maintenance requirements.

Once this data is gathered in a cloud-based system, it can be analysed to identify equipment status, monitor efficiency and detect if components are failing. Achieving this would put an organization on the digital transformation map. However, to get ahead of the competition, they will need to think ahead too. They can’t predict where the next production bottleneck will develop, or which would be the most cost-effective way to organize maintenance for a fault they can’t see – but machines can.

Predictive maintenance is always learning

As more manufacturers embrace IIoT, new opportunities are created for maintenance to move from a traditional reactive to a predictive model and start to have a positive impact on equipment uptime and production quality. With reactive maintenance, if equipment unexpectedly goes offline as a result of failure, the damage to productivity has already been done and creates a knock-on effect further down the production chain. In contrast, predictive maintenance strikes a balance between reactive and excessive maintenance, identifying and resolving potential issues before equipment breaks down without incurring excessive costs from emergency or even over-maintenance.

AI and machine learning represent another step towards truly predictive maintenance. When unleashed on the vast volumes of data captured from the plant floor, data analytics can be enhanced to filter out anomalous information, detect hidden or underlying patterns and more accurately project equipment reliability – and adjust maintenance schedules accordingly.

Data is the lifeblood of digital transformation

Advanced data analytics will be key to identifying any ‘teething problems’ when deploying new, fully digitised equipment and systems. Analysing the effectiveness of emerging technologies and how they affect existing business processes can help future-proof businesses against further digital disruption and manage the impact of newly-deployed solutions.

The benefits of improved data analytics capabilities become immediately apparent across multiple business processes. This was the case when Domino Printing Sciences updated their enterprise systems to accommodate more data-driven business insights. “We are using live data in the manufacturing process to drive improvements in yield through automated test equipment. It removes subjective human assessment and poke-yokes the product at each stage rather than fixing at the end of the process”, said Carl Haycock, UK printer operations director.

Combined with machine learning, data analytics can help improve the availability, uptime and lifetime of assets – helping organizations cut costs, improve operational efficiencies and rely on data-driven decision making. When harnessed correctly, these disruptive technologies will deliver wider business benefits for manufacturers, from shop floor and the customer experience to service delivery models.

By Kevin Bull, Product Strategy Director, Columbus UK