Machine Maintenance Post-Covid-19
Published on : Thursday 06-08-2020
Post-Covid-19, digital will be the new normal for machine maintenance. How prepared is your plant, asks Shekhar Shirwalkar.
Every disaster leaves behind a few learnings and things are never the same again. If I have to pick one after the Covid-19 for the manufacturing sector, it is the imperative of early diagnosis that can lead to timely corrective action. Post-Covid-19 when there is a demand surge and factories have to run above capacity levels, the last thing you want to see is an ‘unplanned shutdown’ due to a breakdown of a critical machine. Imagine the humongous opportunity loss, and you will realise the importance of new digitalised technologies of predictive maintenance. Using smart sensors, advanced early diagnostic, and IIoT enabled remote monitoring through Cloud, these technologies have the potential to eliminate unplanned shutdown. That can be a significant competitive advantage for any manufacturing organisation. The time to digitalise maintenance is now.
The Covid-19 outbreak has affected manufacturing industries like never before. When the situation starts getting better and the business resumes operations, everyone anticipates that the world won’t be the same again. Industry experts are seeing that digitalisation and automation would gain much more significance in all business processes. But there may not be enough clarity as to how one can prepare for the change? Where to begin? What to prioritise?
One area I see as both critical and easier to implement digitalisation is machinery maintenance. Assuming your industry has a demand surge post-Covid-19 and your plant has to respond, the last thing you want is an unplanned shutdown in your production line because one of the critical machines ‘suddenly’ broke down without ‘enough symptoms’. These types of opportunity losses make it imperative to adopt digitalisation in maintenance.
There are misconceptions around digitalisation; such as, it is complicated, expensive and needs large scale. Not true! The key is to identify the right applications for a pilot and then scale-up. According to a Mckinsey study (https://mck.co/2Oi28md) A select group of industry-leading global manufacturing companies are using digital transformation to develop new or enhanced ways of operating their businesses, using a variety of Industry 4.0 capabilities. The benefits these companies have recorded include 30 to 50 per cent reductions of machine downtime, 15 to 30 percent improvements in labour productivity, 10 to 30 per cent increases in throughput, and 10 to 20 per cent decreases in the cost of quality. These breakthroughs create impact across the value chain that may be even more important, if harder to measure – increased flexibility to meet customer demand, faster speed to market, and better integration within the supply chain.
I have recently seen how a few eminent companies in the metal and cement segment have embraced digitalisation for machine maintenance. After the initial consultation and site assessment, they focused on one critical process line to be built as a pilot that is ‘protected’ with advanced digital predictive maintenance technology. Each critical equipment on the line is monitored with appropriate sensors that collect real-time data that is processed in real-time and the machinery health status can be seen in real-time everywhere, from the console in the maintenance workshop to the DCS and remotely on the cell phone of the COO.
There is another widespread belief that the digitalisation of maintenance in a brownfield plant is hard to implement than for a greenfield project. I don’t agree with it completely, thanks to today’s advanced technology options that make things simpler than we can imagine. Most of these technology components are modular, scalable and support plug-n-play. I call this a Lego- style design approach.
To highlight the benefit of modern technology over the contemporary practice, I refer to the use of advanced ‘Torsional Vibration’ analytics to diagnose the problems in rotating machines. Refer to the typical PF curve below. The benefit of being able to detect the fault much earlier with Torsional Vibration compared to the contemporary ‘Linear Vibration’ technology are quite apparent.
Imagine an early warning of an emerging fault on a critical bearing or gearbox. It can be as vital as detecting cancer at stage zero. A timely diagnosis enables timely corrective action thereafter. It will not only prevent the abrupt shutdown of the process but also save that critical equipment and a key component within (like a bearing) from getting fully damaged. You can calculate the losses pertaining to one unplanned shutdown on your production line, add to it the inventory, repair and replacement cost of the troubled machine, manpower and energy costs and it’s easy to imagine that the investment in the digitalised torsional vibration analytics system has a very high RoI.
Another example is oil management. Research shows that more than 54% of the machine problems occur due to not maintaining oil hygiene. There are many parameters like moisture, temperature, particulate matter, oxidation, TAN and TBN values, etc., that determine the oil quality. The machinery OEM, as an example an engine manufacturer, ‘recommends’ an oil change interval. Given your own usage and maintenance, the oil change may be needed earlier or
can be extended. But how do you take a call? You need real-time reliable data that the traditional patch tests or even the lab tests can’t provide. I am aware of how some of the heavy users of diesel and gas engines from the marine and oil & gas segments are using this modern digital technology. Just one single sensor that can account for the impact of all different sources of degradation and gives one simple ‘indicator’ that tells if your oil is good enough for continued use and if yes, how long can you keep using it at the current rate of degradation? The sensor is fully digitalised to give you an on-site display of diagnostic and also connects via IIoT to be integrated onto a comprehensive Asset Reliability Management dashboard. These users have been able to take critical decisions of extending the oil change intervals by up to 3X without risking the reliability of the engine because they had ‘data’ to prove that the oil is in good condition. The savings run into millions of rupees over the operating lifetime and the payback is in less than a year.
From my experience, I can tell that to get the optimal results with a justified RoI it is necessary to deploy an Asset Reliability Management solution in an end-to-end manner across the process line. This is another evolution over the earlier contemporary practice of doing predictive maintenance only at an ‘equipment’ level and using a ‘spot solution’. The end-to-end approach that I recommend is illustrated in the below example where a customised detailed engineering solution has been designed and implemented to enhance the reliability of a ball mill in an aluminium processing plant.
As shown in the schematic diagram, the system uses different online sensors like torsional vibration, oil quality management, pressure, temperature, etc., on various critical machines so as to map their health in real-time (sensors indicated in green and red). This data is processed by edge computing devices in the field and the output is sent using an IIoT platform to the cloud- based dashboard that is visible for all relevant stakeholders, from the maintenance team in the plant to the COO in the head office.
It is also important to design such a system grounds-up. I call it an ‘operator-centric’ approach. The diagnostic displays, analytics, alarms, and the overall dashboard has to be simple enough for a machine operator to understand and take action. Unified indexes and colour-coded status alarms and singular parameters work the best, over fancy and complex curves, waveforms and multiplexed displays that essentially require an ‘expert’ to read and explain.
Knowing very early that the bearing is in the critical alarm stage is more helpful than a detailed analytics report pinpointing exactly what is wrong in that bearing but delivered just before the bearing fails. Time is the essence here!
The time for digitalisation is now and maintenance is the low hanging fruit where the risk to returns ratio is favourable. The investment is moderate too due to the scalability of the solution starting with one production line at a time. The way the world will transform post-Covid-19, digital predictive Asset Reliability Management will be the new normal for machine maintenance. How prepared is your plant?
Shekhar Shirwalkar is Head of Marketing, Neptunus Power Plant Services (https://bit.ly/32bmaa6). He carries over two decades of experience in marketing of technology driven B2B solutions across diverse domains of industrial automation, construction material, renewable energy, pharmaceutical to name a few. Shekhar is currently heading the marketing function at Neptunus Power Plant Services, which is serving customers in industrial, data centre, marine, defence and oil & gas domains with customised engineering services and solutions. Shekhar has published articles in B2B media and also expressed his views as a speaker in the events hosted by CII, UBM and other industry bodies. He has done his Masters in Marketing and Bachelors in Electronics Engineering from Mumbai University
