Six Questions that Guarantee Your Daily Win
Using SPC or Statistical Process Control charting and basic time tracking for downtime/defects
The backstory is that we were in the middle of our ramp up phase in the new plant. Managers, supervisors, and operators alike were all learning the equipment and production processes. We were struggling to improve performance. In a discussion with the team the comment was made – “how do we know what to improve if we don’t know what’s happening?” We had machine data available. We have various HMI screens that give current machine performance. But, as a whole integrated production line, we did not have visibility of how everything worked together.
That gap birthed the shift playbook. It was designed to give the supervisors a current state performance metric by capturing the hour by hour performance, visualizing the previous hour and the shift trend, identifying targets, and tracking abnormal conditions or events that led to downtime.
There are two parts. The first one is a standard SPC chart that displays each hour slot of a twelve hour shift. The minimum shift average of 40,000 parts per hour is designated by a change of color from green to red…simple yet genius
SPC Chart – The First Three Questions
The SPC chart is used to answer the following three questions:
- How did we do?
- How are we doing?
- If nothing changes – where will we end up?
How did we do?
Depending on your operation there can be differing amounts of variation in your production process. Analyzing this particular process it was decided that hour by hour would be the optimal measurement and would allow some variability that occurs through each sixty minute cycle to normalized over time.
At the end of the hour, the supervisor places a dot on their playbook sheet in the closes square that represents the actual production performance number. Plotting the following data would result in the following visual representation:
- 06:00 – 07:00 – 47,521
- 07:00 – 08:00 – 54,596
- 08:00 – 09:00 – 34,234
- 09:00 – 10:00 – 32,335
Answering the question “how did we do?” is easy when represented with a simple SPC chart. We can see that we started off fine, things got better and then something happened that caused productivity to fall below our target minimum.
How are we doing?
We can also answer the question “how are we doing?’ by looking at the number in the shift average row. We see that given the current hourly performance, the shift average is starting to go down. It is easy to see that we are at risk.
If nothing changes – where will we end up?
A short while ago, I was doing a Gemba walk with the on-shift supervisor. We reviewed the Shift Playbook for the current state. It was a similar situation – the dots were pointing in a downward direction. I called an all-hands meeting on the production floor with the supervisors and leads. We looked at the playbook as a group. When I asked the question – “If nothing changes where will we end up by the end of the shift?” Everyone knew the answer. If nothing changes we will either continue to decline or remain below goal for the shift.
That quick answer then led us to the next section of the playbook where we answer the second set of questions.
The Production Delay Log
The production delay log is a simple yet powerful tool when combined with the SPC chart above. It helps answer these questions:
- What is getting in our way?
- What impact is it having?
- What do we need to focus on to improve?
We had decided as a team that just like tracking the hour by hour at a minimum, that there was too much noise and busyness to track every delay. We aligned that we would only track incidents that resulted in delays greater than 10 minutes. Your operation may be mature enough to tighten the interval for tracking but as a startup facility we found it more value added to address the larger items first.
This would also keep us focusing on the critical items and not chasing everything.
Reviewing a sample delay log that would match the SPC chart example above we would see the following:
By keeping a general production delay log the information from different machines or areas of the production line can be brought together to see if there are any patterns.
The data answers the last two questions. 5/6 or 83% of our delay is caused by the occurrence of twisted cases getting to machine 3. We can also see that of the total downtime, approximately 60 minutes – 50 minutes are caused by one type of incident.
It is easy to see the pattern here. When the team reviewed the playbook there was agreement that we need to Gemba machine 3 and see where the problem is coming from.
That is exactly what we did. The whole team went to machine 3 to watch the delay happen. While we were there we saw a twisted case arrive – already twisted. We decided to move upstream. We went to the next machine which was a combiner. It takes two different source conveyors and combines them into one line a moves them to a palletizing machine.
We watched for about ten to fifteen minutes and then it happened. We saw a case get hung up on one of the side / edge rollers that merged the case to the other line. It didn’t happen every time but as we continued to watch we noticed that upon closer inspection, every case looked like it bumped something – it caused the case to shift slightly. The severity of it was random. Every ten to fifteen minutes it was severe enough to cause the case to turn 90 degrees and stop machine 3 downstream.
We called maintenance right away who made a slight adjustment to the angle. We then stayed to validate whether our hypothesis was correct. The “bump” was gone and no repeats of the incident occurred over the next thirty minutes.
The playbook was completed with a much improved hour by hour and upward moving average. By the end of the shift the team finished with the fifth highest shift production record as of that date.
Using and combining the two simple yet effective tools – an SPC chart and production delay log into a Shift Playbook can provide the data and visualization to help your team know what defines winning, understand the current state, and identify roadblocks that are preventing them from winning.