An initial driver of the industrial digital transformation that we’re all living through is production metric tracking. This factor is so intrinsic to industry’s evolution that once all the dust around Industrie 4.0, the Industrial Internet of Things, and other associated initiatives settles at some point in the future, one thing all industrial companies will likely have in place is advanced analytics software to track operating metrics at a highly granular level.
Therefore, it makes sense to start getting a handle on your production metric tracking now to be in a better position to apply this data as it becomes more critical to your company’s operations.
We recently connected with Sam Russem, senior director of smart manufacturing solutions at system integrator Grantek, to gain some insights into current production metric tracking strategies as they develop amid industry’s ongoing digital transformation.
To access critical production metrics, Russem said there are three things you need to do:
- Access the source data;
- Extract the data out of the system or asset in a dependable way; and
- Put the data into context so it can be used by stakeholders to make decisions.
Russem pointed out that MES (manufacturing execution system) software can do each of these things in unique ways.
“When it comes to collecting data, your MES usually sits above your plant's control systems [where] it can talk to PLCs, SCADA, and historians,” he says. “But it's also going to talk to things outside of the factory, such as your ERP and supply chain management systems. MES software is good at getting all of this different data from all of these different systems, and it usually has all the infrastructure behind it to do data computations via database calls, built-in functions, or support for writing your own custom code.”
With these common capabilities, presenting the information via an on-screen dashboard or report is usually the easiest step to getting the data out for decision making.
“So if you're wondering where you can be calculating certain metrics that are tough for you to get today, MES software may well be the block you should be building on,” said Russem.
Tracking the right metrics
If you’re deciding what metrics are best to start of tracking with MES software, Russem said one option is to identify your biggest problem area first. However, this approach can carry some concerns. For example, if you're If you're looking for the biggest problem, you're probably also finding the most complex problem, “because if you could solve it, you would have solved it already,” Russem said. That’s why this approach can lead to a “very complex (MES) design and expensive and risky implementation.”
Another option is to first focus on low-hanging fruit, i.e., obvious, easier to address issues to help prove out the MES software’s capabilities and familiarize yourself with the software before tackling bigger issues. While this approach is a good one to help you learn the software’s capabilities with a low-risk application, Russem cautioned that, while you might be able to show specific results easily with this approach, you might not be as likely to get much direct value out of it either. In other words, this lower risk, easier approach can also deliver a low reward potential.
With these cautions in mind for each of the three approaches, Russem advised that the best approach is “seek a balance between these three. You want to find a problem that people are complaining about and that fixing it is going to provide value—but it shouldn’t be your hardest problem either. Plus, you want to make sure that your approach is aligned to some type of overall business goal that you can set a goal post around. For example, I want to reduce scrap by 10% or increase throughput by 4%.”
With this balanced approach in hand, you can collect all your pain points and opportunities that can drive this metrics approach to which your MES software can then be applied to deliver results.
Assessing analytic options
According to Russem, any MES software “worth its salt today is going to give you descriptive and diagnostic analytics. The question is: Which ones can go beyond that to do more predictive and prescriptive work as well?”
While many predictive maintenance software tools can provide data similar to what MES can provide in the maintenance arena, Russem noted that there are predictive maintenance tools using similar data to an MES that are “purpose-built and very, very good at predicting” asset maintenance metrics.
Using such software can be a strategic aspect of your operational analytics because you might not want to do certain analytics in your MES if there's software that's out there that is designed to specialize in that area.
Russem added that you can also layer artificial intelligence platforms on top of directed analytic software packages if you want to create something even more powerful. He stressed that, even with these added analytic capabilities, we're not talking about taking people out of the equation in most of these cases.
Even if you have a great deal of intelligence in these systems, what they’re really doing is “pointing you in the right direction and telling you what to fix,” Ruseem said. “You still need some human interaction to validate that decision or to actually go through and do the work or make the change.”