This is why you should look into predictive maintenance, right now

Posted on: September 9, 2019 by Stephane Ben Soussan

Time is valuable. This is evident in manufacturing. Availability and utilization are critical concerns in any facility that relies on machinery and industrial equipment. If there are routines, there are efficiencies to gain. Predictive maintenance is an Industry 4.0 IIoT staple and targets what may be the most critical efficiency metric of all: time.

Time matters because equipment downtime and maintenance come with high direct and indirect costs. Lost production, wasted labor, depleted inventory, as well as opportunity cost; these are all results of downtime. Underestimating the amount of downtime is also a costly, and common, mistake.

What is predictive maintenance, and what are the benefits?

How often do you ask yourself "what day was this produced", "with what process" and "with which machine"? How often could the mistake have been avoided if there had been a better overview of equipment condition?

Predictive maintenance in IIoT promises to help you avoid spending time putting out (metaphorical, we hope) fires. With improving sensor technology and connectivity, it has become an Industry 4.0 must-have.

The concept is to use on-premise and cloud computing to widen your window to schedule maintenance. You don’t just assess the condition of industrial equipment centrally and in near real-time - but predictively to forecast issues and machinery life-time.

As a result, you reduce lost production and wasted labor, and you are freeing up valuable time for yourself and others. The potential for optimization-led cost savings promises redirection of maintenance budgets toward research, marketing, and innovation. In short, the potential economic impact is significant. In fact, McKinsey has estimated that predictive maintenance can reduce maintenance costs by 10 to 40 percent.

How does predictive maintenance work in Industry 4.0?

To summarize, predictive maintenance offers a data-driven understanding of failures. Anomaly detection - a machine learning technique - is used to identify failures and predict the likelihood of future failures. Does "Where are we on the Gaussian curve?" sound familiar? If so - good! The normal distribution is one way to go about it.

Based either on static rules or more dynamic approaches, and using data on critical health values, risk predictions, outage costs, and maintenance costs, analytics then trigger alerts for actions to be taken.

The opportunities - Unreaped benefits of investing in cloud and IIoT

Another, so far unrealized, gain of predictive maintenance is that manufacturers will be able to increase the agility of their offerings. Sure, adopting as-a-service and pay-per-X offerings will increase the complexity of production. More importantly, it will also reduce downtime and bring in new income streams. The benefits are enabled by, for example, iPaaS solutions such as dizmo.

As a result of smart factory development, predictive maintenance in Industry 4.0 is part of the sharing-economy trend of increased speed and customization. In the future, manufacturing will become more customer-centric and more service-centric.

Indeed, some managers view Industry 4.0 as a sort of Lean 2.0, and their customers will benefit as production becomes faster and more cost-effective. Now, how do we get there?

Looking ahead

Exploring industry beyond predictive maintenance, futurists are looking toward prescriptive, rather than predictive, maintenance. Prescriptive maintenance aspires for a completely automated approach. It uses machine-to-machine (M2M) techniques to allow machines to schedule maintenance for themselves.

However, manufacturers with expensive equipment might, at least initially, want the option of human input. Thus, to fully capture the value of digital transformation, companies must look end-to-end for technical solutions that are human-compatible and keep up with evolving technologies and business trends.

The roadmap to implement predictive maintenance does not just require appropriate data management. In complex system integration, there is also a need for flexible platforms that pull together the full range of data sources, tools, systems, microservices, and eventually, allow for seamless human-data interaction.

In conclusion, predictive maintenance is transforming manufacturing. One thing is clear: that when investments in cloud technology and IoT come into their full force, they hold the keys to unlock time-efficiencies, but also completely new ways of working.

Make sure to check out some of the uses that customers have made of dizmo’s patented technology. Dizmo predictive maintenance demo for Smart Building is installed in the BTIC (Business Technology & Innovation Center) in Atos HQ in Bezons, and is being integrated into the global smart plant solution from Atos. We'll be discussing this topic and our continued collaboration with Atos at the Atos Expert Conference in Madrid this week.

The partnership between Atos and dizmo started when dizmo won the Digital Industry Award co-organized by Atos and Siemens and lead to several joint activities.

Share this blog article

About Stephane Ben Soussan
Head of Operations at dizmo
Stephane is Head of Operations at dizmo, the Interface of Things. He has more than 20 years of experience in sales, business development, strategic partnership, operations and R&D within high tech environments. He has a large industrial and technical background, made of semiconductors, application SW, integration platforms and consumer electronics. Stephane holds a masters in microelectronics.