Create flexibility in your production schedule without adding a new machine. Use your skilled workforce to their best without any additional training. Sound too good? It’s more simple than you realize. Use the method outlined below to drastically reduce machine downtime!
You have probably heard of, or used the 80/20 Rule, also called Pareto’s Principle. If you haven’t, it’s the “rule-of-thumb” that 80% of the problem is caused by 20% of the causes. Sales & Marketing departments use it to find which 80% of sales / revenue comes from 20% of your clients. It works, and its pretty easy to apply as a problem solving tool. We can use it to determine which 20% of your machines cause 80% of the Downtime, or which 20% of downtime reasons cost you 80% of lost production. Here’s How to do it.
It (the Pareto Analysis) works, and its pretty easy to use to as a problem solving tool.
Step 1 – Decide which machine to investigate.
Start gathering data (the old-fashioned way with paper and pencil.) In the example below we look at why Machine EM5, in the Milling Department can’t produce parts at the same rate as other machines in the department.
Step 2 – Collect data for a period of time.
A week, two weeks, or a month are usually good periods – it all depends on how often the machine runs. Get a clipboard, make people write down (3) things when a machine stops running because of a problem.
- Date and Time of day the Machine stopped
- Date and Time of day the Machine started back up
- Reason the machine stopped producing
Here’s an example…(very rough & ready as they say…)
Step 3 – Group the recorded events into categories.
Machines stop running for many reasons. There maybe material shortage reasons, part quality reasons, or maintenance reasons. Group the recorded events into these categories, to focus on the broad issues.
For this case we tracked total lost time due to each listed cause. You can decide what is the best to report on; total number of events, cost of resources used to restore production, or any other factor critical to your shop and related to this machine. You want a standard measure for comparing between the causes.
For this case we studied total lost time due to each cause.
In the example above we will report on lost time due to each listed cause (i.e. coolant leak, tool breakage, etc…) We logged 33 events over 30 days. Right THERE – you should sayWoW!
That’s a downtime event every day! BUT it gets worse – This shop runs a two-shift operation, and a 5-day work week. So that is 33 events over 20 work-days, or 1 to 2 events every day! We found EM5 was down for 30.6 hours, almost (4) full shifts out of 40. This information alone is worth gold. BUT it gets even better because we tracked WHY it went down.
We logged 33 down-time events over 30 days. Right THERE – you should say WoW!
Step 4 – Create a Chart
Put your pen and paper data into a spreadsheet, or have someone do it for you following the images in this article. {You can also download the spreadsheet used in the article.} Make four columns;
1st = the list of events;
2nd= the total number of each event;
3rd = total number of minutes of downtime for each event category;
4th = the percentage of time each event category is of the total.
Step 5 – Rank the Events in Order
List the items from the most frequent to the least frequent. The cumulative percent for an item is the sum of that item’s percent of the total and that of all the other items that come before it.
Step 6 – Make a chart of your Data
Use a spreadsheet program to graph your data 9the total # of minutes of downtime for each category, and the percentage of time each event category is of the total. Plot each on a separate axis. See the example below. You can do this with graph paper and pencil, it jsut might take a little longer.
Step 7 – Analyze your data!
In this example the first four categories comprise 80% of the downtime. These are: Quallity Issues, Material Outages, “2nd shift can’t find the part”, and Coolant leaks. Fixing these four types of events will result in an uptime gain of 1,500 minutes of additional production (25 hours.)
At an average machine cost of $80USD/hour, you just found $2,000 per month on just ONE machine! Congratulations!
Your NEXT STEP:
Fishbone Diagram – The Five Y’s and Path to Solutions! (Link internally)
Sources: https://trumpexcel.com/dynamic-pareto-chart-in-excel/
