Not everyone has the luxury of redesigning the whole plant.
If tomorrow, your CEO gave you an unlimited budget and carte blanche to change anything in your plant, how would you react? Would you be overwhelmed by all the possibilities? Would you suggest only minor improvements, confident you already run the smartest, most precise plant going? Or would you pull out blueprints you’ve been poring over since your first day on the job? We hope it’s the latter, because at least one of your competitors also has blueprints prepared for updating their manufacturing plant, or will very shortly.
Let’s rephrase our earlier hypothetical. Suppose the CEO offered you the opportunity to redo the entire plant with an unlimited budget, but on one condition: You need to provide one unifying principle for running the plant, one word to describe your manufacturing philosophy. We know what we would choose: simplicity.
We had to do that but this is a serious article. Really.
Think KISS, or, Keep It Simple, Sweetie/Smarty. This acronym has a long history, with precursors in Occam’s razor (“The simplest solution is most likely the right one”), playwright Clare Boothe Luce (“Simplicity is the ultimate sophistication”), and Shakespeare (“Brevity is the soul of wit”).
The point is, given the opportunity to streamline your manufacturing data operation, who wouldn’t want to make things easier to run, track, and change?
Your CEO isn’t likely to offer you enough money for a complete plant makeover, of course. Expensive capital replacements might not seem high-priority if your organization is doing “well enough.” Inertia is hard to overcome. There may be a notion that any corporate inefficiencies rest with another department likes sales, marketing, or even HR. At best, you’re probably used to receiving funding, when you do receive it, to make only small changes.
While it seem necessary to burn everything down (figuratively) and start again to get the results you want, there’s another option: executing a full redo digitally. This would cost much less money than a material revamp; just as importantly, it would also be faster and simpler to both plan and complete. We believe simplifying your data and its supporting infrastructure is more prudent but just as productive as starting from scratch.
Specifically, we recommend:
It may seem obvious that reviewing superfluous data is unproductive. But how do you determine either data usefulness or uselessness in the first place?
The trick is to clearly differentiate between collecting data and using data. Collecting data is cheap and getting cheaper every day. Over the last half century, data costs have declined by 30% annually. Collecting data will only become cheaper. Because of this, companies naturally accumulate more and more data: more sensor data, more forms, more spreadsheets.
In a world with increasingly vast amounts of data being continuously produced and collected, knowing which data to spend time analyzing becomes a similarly crucial, in-demand skill. At 3AG, we determine the most useful, accurate and profitable data to examine by integrating statistical analysis into all our core services. We apply correlation analysis to more effectively model digital operations—and can build AI and ML models in a fraction of the time it would take otherwise.
This process can be used to build executive dashboards and generate useful reports, built as it is on the recognition that not every KPI is worth measuring, especially if it isn’t a leading indicator for specific business requirements and goals. It makes good sense to collect a lot of data but it doesn’t follow that you need to use it all, or use it all now; intelligently modelled data operations make this distinction for you.
Collecting a lot of data can also introduce errors into your system. Bad sensors are one potential source of bad data, but they are often the easiest to identify (by tracking a reasonable range of sensor data values; if data is outside this range, alarms can be set).
More insidious are the minor errors that crop up when people manually enter data. The more steps completed manually, the greater the likelihood errors will be recorded and disseminated; this happens when staff copy data from one source to another, say from notebook to spreadsheet, spreadsheet to spreadsheet, or spreadsheet to PowerPoint. We’ve seen this sort of data disaster firsthand, in a billion-dollar company no less: Data transferred from notebook to spreadsheet, to pdf, to spreadsheet, and finally to PowerPoint!
As a general rule, data that can be sensed, extracted, or pulled automatically will not be error-prone. So, why doesn’t everyone make the effort to automate data extraction? Primarily because of the false perception that it’s difficult to automate data extraction.
Data engineering is designed to find ways to automate data collection and archiving; it includes an amazingly diverse set of technology options for working with a wide range of data sources. Data warehouses can be used to automatically pull data from spreadsheets and old VAX systems, to modern ERPs and CRMs.
Automating corporate data collection and analysis reduces the likelihood of errors multiplying and skewing key business information; it also streamlines processes. Perhaps most importantly, it can provide the most useful picture of your data possible.
This leads us to our final suggestion: Simplify your reporting. Just because it’s possible to report on every possible metric doesn’t mean you should. In many cases, superfluous charts obscure the real insights a report author is trying to convey. Such unnecessary content can cause distraction, and, in a worst-case scenario, lead stakeholders towards bad conclusions.
Many self-service BI platforms enable users with little or no experience to create reports and charts. This is great for employees exploring, but it can become disastrous if everyone at a company begins pulling reports and promoting their version of the truth as the only truth.
If you have the advantage of virtually unlimited data collection, and you know which measurements are the leading indicators with the strongest operational correlations, finish strong. In this context, finishing strong means developing dashboards and reports that properly communicate a handful of key insights about and useful to your organization.
Self-service tools aren’t always a bad idea; they can be used as powerful platforms that support straightforward reporting centralization. Data can be automatically pulled into these reports, which means you can trust its accuracy (having done the work to automate collection in the first place). This means you can limit reporting to a small set of dashboards and reports everyone in your organization can access—but not fiddle with—from a single source.
Simplifying your reporting is a crucial part of refurbishing your data infrastructure so it’s accurate, insightful, and secure.
Keeping things as simple as possible is critical for manufacturing success. Applying the same philosophy to data management and reporting is equally critical.
Just like your equipment, your data system should be one of your company’s strategic advantages. It should allow you to collect readings on all operational functions, but also make it clear which readings are worth looking at. It should allow employees to focus on the most important issues at hand. And it should keep everyone on the same page, working together to achieve shared corporate priorities.
At 3AG, we simplify complex operations, guided by a relentless focus on the “KISS” principle expertly applied. We help organizations like yours do much more and do it smarter—with a lot less money, time, resources, and effort.
We just sent you an email. Please click the link in the email to confirm your subscription!