"OEE" looking in a different angle
Background...
During my last years of working on Industrial Automation domain and with various Industrial manufacturing customers there is some common requirement from all, that is everyone wants to improve their performance, quality and availability of their productions machines to the maximum.
Availability of the machine for production Is very much required for the production process, the less the rejected count of the product the more is the production value.
Performance improvement is when we identify the causes of the problems and fix or improve them on time.
To achieve a better efficiency of the production process all 3Ms has to be looked into i.e. Man, Material and Machines.
Most of the MES applications used for this purpose and there are also custom applications developed to monitor and improve these KPIs.
Some of the KPIs monitored in a typical production environments are listed below.
Out of all these KPIs, OEE is one of the KPI which is very important and in some way can improve all other KPIs if improved.
So let's try to understand this(OEE) in our discussion today.
Let's discuss in details...
So the basic formula for OEE calculation is
OEE = Availability ∗ Performance ∗ Quality (A*P*Q)
You get this if you google it or ask someone, so what is the point to discuss again if it is available on public domain ?
We will not discuss what OEE is, we will focus on some points why OEE is important and what insights a customer/user gets out of it.
For me or for any of the user i think it is important to know what value or improvements we can achieve by knowing this KPI. What internal details it can give us so that someone can take right action at right time to improve the efficiency and performance.
Availability = Total run time of the machine / Production time planned
That means if my machine has run for 6hours and i have planned for 8 hours then, the availability of my machine is 6 hours / 8 hours which is 0.75 which we can also say as the machine's availability is 75% of time.
This indicates that the machine is not able to produce for rest of the 25% of time due to some reasons. This directly indicates there might be machine breakdowns, unavailability of man power for production, unavailability of material, timely feeding of raw material and many more issues.
We need bring out those reasons for the user so that he knows what is the actual cause for the lower availability of the machine.
2. Now let's talk about the 2nd parameter "Performance" and what insights it can provide to the user.
Performance of the machine is basically related to the speed loss or a comparison between the current running speed and the maximum capability(running speed) of the machine. In simple terms my machine can produce 100 parts but it is producing 80 parts due to some reasons.
Performance = (Ideal Cycle Time of the Machine ∗ Total Parts produced count) / (Machine Run Time)
From the above calculation for performance we can find some insights like what is the parts produced , what is the planned production, what is my machine run time and planned runtime. If the machine is not running as per the planned time then why ? If we provide these insights to the user it will help them instead of just showing a % value for performance.
So what are the biggest reasons(loss) for the lower performance, lets have a look.
Machine Downtime Losses:
The breakdown of the machine belongs to the downtime
losses because machine failure will result in the product
quantity losses.
2. Set up and adjustment time loss:
The set up and adjustment means the losses because of
the breakdown or faulty products that were produced
when the equipment stopped the on-going assembly and
adjusted or set up itself to manufacture another item.
Speed Losses:
3. Idling and minor stoppage:
The production process might be disturbed by an
accident, human error, or if the
equipment is idle.
Recommended by LinkedIn
4. Reduction in speed:
Losses occur is the speed of the machine is slower than
planned.
Quality Losses:
5. Defects and reworks of the produced parts
This refers to the bad parts or faulty parts produced in the machine due to
faulty process, operator inefficiency , faulty machine etc.
6. Start-up losses:
Start-up refers to powering the machine on for production.(The time it takes to fully function and ready for production from a switched off state).
3. The last parameter "Quality" and what insights it can provide to the user.
Quality = Produced good parts / Produced total parts
Produced good parts = Total count of parts produced - Total Bad parts produced
The above calculation gives the quality of the parts produced on a particular machine.
This parameter basically defines the FPY(First Pass Yield) of the machine.
The higher is the good parts produced the higher is the FPY.
If quality improves then basically your process , operator efficiency has been improved.
So while designing and developing applications as SW developers we should look all these aspects of data and provide the insights in the reports so that the user can analyze details of it and come up with a improvement plan.
Let's have a look at the below diagram to understand what data is associated and required to derive an OEE for a machine.
If you analyze the above ER diagram you see there is a worker which can operate on one or many machines. The operator has data associated with him/her like ID, Name, his shift time and personal breaks.
A machine has a unique ID, can produce parts, has downtime and runtime data associated with it.
When we talk about downtime we also need to capture those reasons for downtime. The downtime has a reason, type of that reason(Machine failure, Material unavailable, Operator break etc.) and for which machine the downtime occurred, these are the data associated wit it.
Machine can produce part and the Part has count, good or bad part, Part ID data associated with it.
I discussed this entities because while we develop an OEE or any application related to production KPI we need at least these data points to be captured.
Hope this gives you some insights.
End Note:
In the time of Industry 4.0 when many manufactures are moving towards digitalization and many of them are also very new to these production KPIs, I feel it is important for them to know not just the number. They should find out and understand why the OEE is low or OK.
What are the causes for a lower OEE, what are the reasons for different losses ?
What are the areas need to be looked for improvement, if they need to improve their process or the operator efficiency or the supply of material, quality of the raw materials, or timely maintenance of the machines etc.
Excellent