Downtime Tracking and Data Bias

Data Presentation Bias

You’ve implemented your new ERP with your state-of-the-art MES, and now you’re tracking problems on the shop floor!  The MES has some cool reporting features, like built-in Pareto charts and drill-down capability.

After a few months, you check in with the Plant Manager.  Are they using the new systems to get a better handle on issues?  Well…it turns out the reports aren’t that helpful.  Operators don’t like the system.  The reports are full of charts and numbers, but the data isn’t meaningful to what’s happening day-to-day.

You know other companies use this system, and the word in the industry is that it’s really powerful.  Trusted associates swear by the functionality.  What could have gone wrong?  I’ll tell you the three major areas where manufacturing companies fall down.

Downtime Tracking in MES

Before I dive into the specific failure points of tracking downtime in MES systems, let’s define MES and the characteristics required to accurately track downtime.

MES is an abbreviation for Manufacturing Execution System.  In a manufacturing industry such as roll forming, you must have a system that is directly tied into the operation of the machine.  Thin client stations installed at roll forming lines are mostly a waste of money.  These can be useful for manual operations, but not for automated continuous operations.

Will you collect some data?  Sure, but you aren’t going to get accurate data in terms of scrap and downtime unless the MES station is electrically integrated into the systems that run the machine.  There are some options out there to do this –

  • You can spend a lot of money to build your own, homegrown setup.
  • You could work with a third-party industrial controls integrator to come up with something.  That’s also going to be very expensive.
  • You can install an in-between solution.  FreePoint Technologies has a very generalized system, but it’s probably best for manual processes, such as Press Brakes and Cutdown Saws.
  • You might already be using an off-the-shelf MES solution that includes a machine controller if you’re in the roll forming industry.  For fully-automated roll forming lines AMS Controls or Beck Automation are the major names in the industry.

The reason that most roll forming companies go with an off-the-shelf solution such as Beck or AMS is that these pre-packaged solutions are measuring both time and material right at the point where both are consumed – at the machine.  Electrically, these systems should be in direct control of the machine functions, such as Run Mode and operation of the cutting tool.  Most importantly, the machine controller provides software enforcement of Operator procedures.  If your system doesn’t do this, then the data flowing back from it is useless in terms of tracking scrap or downtime.  Garbage data flowing into reports means garbage data flows out from reports.

Common Areas for Failure

The three most common failure points in downtime tracking for MES systems can be found at each step of the downtime collection and reporting process –

  • Downtime Reason Definitions
  • Downtime Data Acquisition from Operators
  • Management Response to Downtime Reporting

Downtime Reason Definitions

When a Plant Manager steps onto the shop floor, there are a million places where he could spend his very limited time, money, and energy improving operations – but where to start?  What’s the priority?

A downtime report with solid data can point you straight at the biggest cause of problems on your shop floor.  It can be used to justify spending money to fix a problem, or even to hire more workers.

The Problem with Definitions

You create a handful of downtime reason definitions, and when the machine stops, the system automatically captures all the downtime.  It then asks the Operator to tell it “Why were we down?”  This can be so powerful, that there’s a tendency in many organizations to try to use the report to basically manage the floor for you.

I have personally seen companies create dozens of downtime codes for every machine.  The thinking is that if the company can capture downtime to a very detailed level, it can drive reporting and cost reasons.  But human beings don’t want to scroll through dozens of reasons.  The Operator should spend his time keeping the machine running, not scrolling through reason codes – trying to figure out if his specific situation relates more to “Cutoff Maintenance” or “Die Sharpening”.

The Solution for Definitions

The research I’ve seen indicates you should keep your reason codes down to 12 – 15 codes.  Create a more general set of codes that easily allow the Operator to make a quick and easy classification for downtime.  Then, if the plant leadership team needs to dial down into a specific issue, they should create a code specific to that problem.  Once the problem is resolved, remove the reason code from the core list.

Downtime Data Acquisition

The best method for collecting data from Operators is via barcode scan and the reason for this is not what you’re thinking.  If you’re using an off-the-shelf MES system, then it probably supports downtime reason entry via barcode scanner.  Many plant leadership teams fail to implement this data entry method for Operators, because they wrongly assume the point is to simplify data entry to avoid mistakes.  Then, they see the built-in interface on the machine controller and think, “It’s almost impossible to mess this up!”  But it is the very nature of the built-in interface that creates problems with getting accurate downtimes assigned for the right reasons – it also encourages missed downtime transitions.

The Problem with Data Acquisition

If you’re using an AMS or Beck, then the controller software forces the Operator to enter a downtime code when the machine has been down.  When the Operator tries to hit the Run button, the controller refuses to enter Run Mode, and instead prompts the Operator for a downtime reason code.  This is a huge issue.

Operators take a machine down for a variety of reasons – shift break, maintenance breakdown, material threading, quality audit, etc.  Once a machine is down for one reason, it’s common for it to stay down for a variety of reasons!  A great example:

The Operator is running an order and shift break occurs right at the point where the last coil is being completely consumed.  The Operator goes on break, but when he comes back he’s going to begin a coil change.  So, he isn’t going to press the Run button to get the prompt from the system to enter his downtime reason.  Now, imagine that a Supervisor breezes by to say, “Safety Meeting!”

When the Operator gets back from the Safety Meeting, he still has to thread his coil.  After an an hour and forty five minutes of downtime, he’s ready to hit Run, at which point he can only select one reason code for all this downtime.  Decisions…decisions.

The Solution for Data Acquisition

Most people are unaware that you can navigate to a sub-menu on the machine controller to force the prompt for downtime whenever desired.  Operators should be trained to enter a downtime code whenever they change tasks, regardless of whether they’re trying to enter Run Mode or not.

More importantly, the downtime reason codes can be printed in barcode format and posted at the Operator Station for the machine.  The same barcode scanner they use to scan inventory coil numbers into the system is used to enter the downtime scans.  The operator doesn’t need to navigate to a sub-menu, he just picks up the scanner and scans the reason code.

When the machine stops running for any reason, a timer starts in the controller.  Each time a downtime reason code is scanned, the time from the last run or the last scan is now counted as downtime for the reason code scanned.  That means if you teach the Operator to scan when switching tasks, he can come back from shift break and scan that reason code *beep* and then begin his coil change.  If the Supervisor walks by to alert everyone to a Safety Meeting, the Operator scans the coil change reason code *beep* before walking up to the office for the meeting.  The partial coil change time is collected before walking away from the machine.  Upon returning from the meeting, he scans the Safety Meeting reason code *beep* and continues on with the coil change.  This is the correct way to get extremely accurate downtime data from roll forming lines.

Management Response to Downtime Reports

For roll formed manufacturing companies, accurate downtime reporting is the Key to the Kingdom in terms of Continuous Improvement in terms of capacity and efficiency.  It’s the area where you have the greatest potential to impact margins and increase revenue without CapEx projects for new equipment.

When I teach plant leadership to read the reports and basic Lean techniques for correcting issues, I can see patterns in the reporting data and tell you with 90% certainty that the problem is, Every time the Operator does X, he’s using method A when he should be using method B, and then you’ll get Y result instead.  But I always recommend physically putting your eyes on the issue to confirm my assumption.

The Problem with Management by Report

There is a tremendous pull to manage by report without direct confirmation.  But even if you assume the best of your Operators’ intentions, you could end up spending money and missing some contributing factor to the problem if you don’t first confirm the specific reason for the problem.  You’ll spend money only to partially fix the issue or or potentially exacerbate the problem.

The Solution to Management by Report

Whether we rely on a competent Shift Supervisor or the Plant Manager is personally checking on the situation, you must always put your eyes and brain on a problem you think you see in a report.  When I do consulting or training, the company always learns something about its people and processes when we print off a report and physically walk out to a machine and watch the Operator for a short period.  In nearly every instance, the Operator is doing his level best with a bad situation, and a few simple,  inexpensive steps on the part of leadership would make every Operator’s life simpler and easier on a day-to-day basis.

Get Help Now

When the plant leadership walks out onto the shop floor, having a powerful downtime report is like having a burning spotlight shining on your biggest issue.  That issue becomes impossible to ignore when you know that you spend an extra 10 minutes per coil change – with an average of 10 coil changes a shift, 2 shifts a day, 5 days a week – you waste an average of 1,000 minutes on that problem.  That’s two full shifts per week of lost production – because of a single problem.

It becomes a simple matter to justify an expensive solution based on the ROI and increased output.  From the perspective of Operations, it also shifts the responsibility onto Sales, because now Operations has just increased capacity in a big way without having to install an entire production line.  It’s the difference between $60K for a new uncoiler with increased capability and $600K (or more!) for new line.

Whether you need to spec out the right MES solution for your plant or you’ve tried implementing one and you’re struggling, I can help.  I spent over 20 years doing this for dozens of companies in North America and around the globe – Australia, Canada, China, Germany, Italy, Mexico, South Africa, and Sweden.

I’m happy to be on-site, but many companies can save money by doing the work in-house while I provide support via phone, e-mail, and videoconference.  Contact me today to schedule a free consultation.  I’m happy to tell you exactly how I can help and how you can get more out of what you’re company has already spent so much money to install.

Leave a Reply

Your email address will not be published. Required fields are marked *