Dear Diary, help me understand what I worked on 🪚
today I am getting back to you, as the Production Guy within our company, in a mood of excitement.
Remember last time I was opening my heart when I spoke about that expensive system, and that the guys from SAP came up with this tool called Process Insights?
Could you imagine my CFO found the diary, and put down his thoughts?
He put pressure on me, but he also ‘kicked something loose’ within me.
In a moment of experimentation, I took the dataset, and worked some magic on it 🤓.
I wanted to answer a pretty simple question:
- Do we finish early production when we start early?
- Do we finish late when we start late?
- When do we finish when we ignore system suggested data
Here is what I came up with (I confess – again I was using on of the things that this SAP SIGNAVIO brings – it is called ‘Process Mining’ ( no, we are not digging holes into the ground with shovels – we are examining Process Data. We have fun giving Names to some of the steps, I was told to call these things ‘EVENT’ ( I like event, such as Concerts – but here we have other events such as : Buying a ticket for the event, travelling to the concert, passing security, enjoying the show, leaving the venue, taking the train, going to bed)…
Now this is what I created:
Let me tell you my take aways before I forget:
- Our system tells us to produce something ( MRP ) – well we act quick, and create the orders even earlier then required – but – we are just hitting the date ( I think we are working with wrong production time assumptions
- When we decide to produce something ( I called this MANUAL ), of course we create ‘On time’ ( there was no proposal), but also here we release the order to production early – argh – but can you see – we sometime deliver late
I know is a bit silly to ask a diary for help, but last time you somehow contacted my CFO – so can you help me again? What should I do?
Dear beloved Production Guy,
First of all, thanks for asking - I appreciate this very much.
I am amazed - both positive and curious.
Based on the data of that tool using Process Mining, I see that most cases try to go via "machine" (MRP), however, it seems lots of those cases do fork down to "man(ual)". Given the fact, that most MRP related production orders go to early creation, early release, but NOT on time delivery, I am worried about inefficient usage of resources.
In a perfect world I would have hoped:
Most if not all go via the MRP, then on time creation and on time release, thus on time delivery.
It seems that manual interactions yield better business outcomes - but only in the beginning!?! From Manual planned order to production order release we have on time creation, but then STILL an EARLY release. Please schedule a session on "the net working capital impact of semifinished products as work in progress on the shopfloor" -
Spoiler Alert: Millions of EURs every month. But we do not shoot the messenger, but appreciate your candor. Like Cobots (collaborative robots) on your shopfloor, I believe in the symbiosis of "man and machine" (actually, Man and Man who created machines), obviously Man being a short form for Human and not male.
Do you know which VARIANT gives me the most headaches, even if it does not happen that often - but once is already once too many!
Manual -> on time creation -> Early Release -> Late Delivery
And of course
MRP -> Early creation -> Early Release -> Late Delivery
I am sure your nice little tool can also ringfence these variants and events separately, it would be very interesting to find out the root causes here. My guess is, that we probably have a "white negative" and/or "false positives" where the data shows the right correlation, but that the causality lie in the difference / mismatch to reality. Meaning, we do know that for example planned order to production order conversion + routings would equal to 3 days, but the BOM (Bill of Materials) require some parts that are not available and thus the system would show a certain delay.
Obviously your colleagues are extremely customer focused and usually do save the day (and get some material from wherever I don't know or sure) and thus deliver on time (based on customer expectations) but late or early based on the master data in our system. Is this a possibility?
On a totally different note - I am also open to the fact that maybe this result is also due to our financial and controlling processes. Do you remember the BAB = Betriebsabrechnungsbogen?
We have not updated our Kalkulationsschema (costing sheet) into the "Regulativ" (the allocation keys based on planned capacity, planned consumption for each work center, cost center, material consumption, etc.) for some time, so maybe it is my team which actually hampers you from doing the work correctly.
In any case, it is mesmerising to see this level of interactivity on processes and I am very eager to learn more.
Looking forward to you and your team to save the day (once again and as usual!).
Keep up the good work!
I am speachless again.
I agree ( of course, how could I not ) to your 'biggest worry variants' - starting early and finishing late.
I have to work with the guys, but please keep in mind, the chart you saw is aggregated data accross all of our plants in Turkey, India, Slowenia, Germany, Croatia and Chile.
So it would not be fair to take judgements and derive conclusions - but we may want to break it down to the different plants and benchmark them amongst each other.
I have seen that the plant in Chile is showing our desired behaviour in the vast majority of cases.
We will trigger an investigation to understand how we can leverage their way of running production so that we improve the other units.
The BAB - I am afraid - I think I am using this one (number 8) when I go to work ( It is called Bundes Autobahn, correct? 🤨).
I have a friend who is deeply into controlling, I will work with him next time he is around, and get back to you in a separate blog post, OK?
Thanks for taking the time - and thanks for having me (still) in your company even if there is so much to improve,
sincerely, your Production Guy