Downtime is one of the biggest problems in manufacturing. It is costly and damaging to all manufacturing businesses.
Downtime is caused by planned maintenance, tool breaks, adjustments and even bathroom breaks. Every minute a manufacturer’s machine isn’t up and running, there is a chance that revenue is lost. What’s more, unplanned downtime can be debilitating to manufacturers that rely on performance and quality to stay competitive in their niche.
Understanding what disrupts machine uptime is critical to preventing machine downtime and improving lean manufacturing processes.
All three of these things are disrupted when machine downtime occurs. Uptime and downtime metrics are actually fairly arbitrary (we talked about this before) if the data isn’t accurate. What’s actually is important is the improvement of these actual circumstances.
What do I mean?
Manufacturers should be more focused on machine downtime than whatever they believe are acceptable metrics for these sorts of events. Achieving industry acceptable OEE or machine uptime metrics mean nothing if you are losing thousands of dollars a day due to machine downtime.
How is availability affected by downtime?
Availability is a measurement of the time the machine is running compared to the planned schedule. You have to measure it to accurately to reduce downtime.
To determine the Availability of a machine or cell you need to collect data directly from the controls or PLCs on the equipment. However, this typically isn’t enough on its own
Manufacturers must also augment this information with input from operators or supervisors to gather context around specific events.
Without human input, you may not understand why shift changeover is taking 15 minutes longer than planned. If all you know is that machines are going down 15 minutes longer than the schedule has planned, how will you ever troubleshoot that issue without that missing piece of human data?
Answering these questions, gaining this context, and solving these problems is the first step towards eventually automating these things (something we have talked about before as well).
What are the top causes of downtime and what are top downtime trends?
Context is key to improving machine downtime. Understanding when the machines are up and when they are down is really important; however, whats more important is being able to understand why they are up and down in the context of history.
We can dive further into the details, looking at individual downtime events to see how long they lasted, if we have the right data to accompany these actual events. This is where that human data comes into play.
If we have inputs from all the parties experiencing the downtime (the people on the floor), we can have much better context as to why it is occurring.
For example: if downtime is 3x as high during 3rd shift, we will want to know why. Perhaps we find out that it is because of common issues like:
or personal breaks
These events are common, but if they are out of line with planning or other shifts, we can fix this.
Manufacturers need to track trends in data that allow them to see how the plant is running over time. Using historical trends, you can easily see the effect of process or system changes over the long term. You can also notice when downtime is increasing or decreasing and quickly take action to resolve the issues.
Stop guessing, start knowing
The top problem that manufacturers have with downtime is what I like to call the conspiracy theory model. They have metrics that aren’t actually tied to objective data straight from the machine and they blend that with platitudes and subjective data from different parties within the business to get a fairly inaccurate picture of what is really going on.
What is needed is a more objective historical view of what’s going on (data right from the machines), augmented with the reality of what happened in those moments (straight from the plant floor) that can help provide real context to the negative events.
This what will ultimately give manufacturers the ability to move away from conspiracies and towards a more data-driven business.
SensrTrx is a really easy way to do this. In fact, we’ve never had a customer stop using SensrTrx after buying it. How about that!