A bearing with 40,000 hours of life runs for under 10,000 hours, and it is replaced to avoid the likelihood of failure. Let's take the first step towards digital twins by making better maintenance recommendations.
Intelligent maintenance is a set of rules that maintenance engineers and experts have learnt the hard way in the field over decades, and retired after their service walking away with that knowledge. Yes, there is some amount of documentation, written in their dairies, or in the form of reports for posterity, but for the most part, it is subject matter expertise that is reinvented with every generation of employees joining the workforce. And as the number of folks in field jobs decrease, the collective knowledge and it's retention decreases leading to increase in downtimes, more expensive repairs, and more frequent outages.
One of our customers is currently getting out of this situation by digitising maintenance procedures for tricky problems like replacement of bearings before end of it's operational life. There's little visibility or understanding of why the bearing overheats causing the side channel blower to stop operations halting the entire manufacturing process despite using heat resistant bearings with upto 40,000 hours of operational life under mild conditions. Since there's consistent breakdown after 90 days, they keep a 90-day window for bearing replacement to avoid an expensive downtime. The same pump has another set of bearings, work well for the intended duration of it's operational life.
Digitise the standard way humans operate - by process of elimination. Our solution captures the current temperature through a non-invasive method that's fed into an edge machine and that continually monitors trend lines to make recommendations. The trick is capturing the right rules for that specific pump operating under specific conditions, and scaling that to all other machines that operate under different conditions thereby tweaking the rules slightly to ensure it's predictions are correct. This is done through Fabrik's flagship no-code editor that allows users to control rules for alerts at a minute level enabling scale across the facility, one pump at a time.
This implementation is especially valuable due to the tight hardware-software coupling between sensors and Fabrik application. The tight coupling enables us to spin-up quick digital twins that truly represent the deployed asset.
Digitising tacit knowledge is the key success metric in this implementation. The digital tacit knowledge creates prescriptive maintenance for components for just-in-time repair and replacement instead of a fixed schedule. Incorporating this feature across the factory floor has optimised the supply-chain and helped plan procurement more efficiently.
From defence to natural resources, and factory floors to public infrastructure, operational twins are adopted at all levels reflecting the strategic benefits of ubiquitous computing devices. Operational twins are the next stage to unlock efficiencies that were previously not possible due to the distributed nature of knowledge. At the same time, due to the critical nature of operations, care must be taken to ensure the single source of truth does not become a single point of failure. This creates a large opportunity for secure information storage and dissemination, along with rapid lightweight computing capabilities for quick decision making.
