Jan 27, 2021
While talking about being data-savvy about industrial as something ground-brakingly new is funny at best, there are still considerable challenges to it. First let’s look at ways to actually gather the right kind of data.
You just use what means are available to you and check the equipment yourself or have your people do it. While human contact is usually irreplaceable to get to know the condition of your machinery, the data will be prone to a high degree of very human error. Starting from basic skill level to gauge the condition of machines to simply being conscientious about it will impact the quality of your insights. In the end you can’t really avoid it, but relying just on the ol’ eyeball wouldn’t be too smart.
Sometimes old machinery just won’t die and will stay in your fleet for decades. The thought that you’re relying on prehistoric generators doesn’t exactly make you happy, but they get the job done and the clients are happy. Not their fault for not being the latest shiny toy with all the bells and whistles, but you’d like to know how they’re doing.
For one you can set up a sensor receiver that can listen for indicator signals and let you know when a warning lights up. Fuel measurement probes will even tell you when to top off the tanks and a vibration sensor will count machine-hours for everything that vibrates when on. The options are endless, but most of them will require a custom solution at the very least on the software side.
Newer machines like generators, pumps, lightmasts and others usually have some way of reading basic stuff like RPM’s, fuel levels, fluid temperatures and let their operators know about them on a dashboard. Nowadays a lot of newer equipment can also come with manufacturer provided remote telemetry software which lets you get and analyze all kinds of data remotely. It has its advantages and drawbacks. Especially if your fleet consists of machinery from different makers.
A way to herd all of your machinery under a single system can be done by implementing a device that reads these data lines and sends you the information via network connection. Usually machinery talks to their control hardware via CAN bus or Modbus data lines. It’s hard to directly compare both, so suffice to say that either can provide a plethora of information. The main difference between the two is readability or rather the way a third party monitoring software can interpret the data. To successfully get data from industrial machinery CAN bus one needs to know what they mean. Usually manufacturers aren’t too keen to give out their secrets which limits the utility of CAN bus for telemetry. On the other hand Modbus has always been an industry standard interface for reading telemetry data which is usually what EngineReader uses. Our hardware can work with CAN lines as well, but mostly this scenario is reserved for machines with no Modbus interface like smaller generators.
The big pie in the sky of course is true predictive maintenance. Ever met a mechanic who can tell right away what’s wrong with a car just by listening to it? Turns out AI can do it as well. Modern vibration sensors have enough bandwidth to “listen” how a mechanism is vibrating and by observing enough failures, it can train itself to hear patterns which lead to mechanical failure. Our RnD people are busy developing this tech which comes with its own set of challenges, but also holds a promise of unheard of downtime reduction.
In the end the data in one form or another will be “eyeballed” by you and your staff, so the human factor remains. What one can do is avoid burdening their staff with multiple softwares because learning them all and getting timely support for them isn’t realistic. And oftentimes it’s prohibitively expensive to just pay for all the different systems. The single goal for the EngineReader team is to perfect our platform which can read everything there is, so companies can work smarter and bring all of their equipment under one smart umbrella.
Let's have a chat - we can help you to level up your business!