Share article:
There is a a famous saying that ‘failing to plan is planning to fail’ and there is a distinct lack of planning and scheduling in manufacturing, and more so in food manufacturing.
One of Chasm’s guiding principles is that Planning & Scheduling are the centre of the universe and all other manufacturing disciplines are a support to planning and scheduling. Operations’ job is to execute the plan that they are given, on time, in full, to the right standards.
However it’s not that simple….
“Our people on the floor know the best sequence, the best machines and how to deliver, so if we just tell them what we need at the end of the day we can leave them to it, right”?
This would be OK if it were true, but alas you can bet your a**e that every shift in a food manufacturing factory has a different idea of what the right way is.
Lets take a deeper look at just how vital planning is-:
In FMCG & particularly for short shelf-life foods planning and scheduling is complicated for a few reasons but first and foremost we are making to a forecast before we get an actual order.
Before we go any further here are a few key terms:
SKU – Stock keeping unit, the unit we sell to a supermarket. Everything with a barcode is a SKU
Sales forecast – is the master file that the commercial team use to ‘guess’ what they are going to sell. In the ideal world the forecast would be 3-way by: customer, SKU, day. In reality, it is usually a financial number that the supply chain team need to convert into something meaningful. The sales forecast is usually wildly inaccurate, if I see anywhere above 70% accuracy it is a bonus. There is usually an ongoing debate over how to measure that sales forecast.: The commercial team will argue that a weekly number is close enough; site would prefer a 3-way forecast (as I will explain)
Plan – this is the piece of work that converts the forecast into a rough plan for the factory, what SKUs to make on what days and how many. Most experienced supply chain people will use historic performance and other known factors (weather, holidays, events) as a guide to daily splits and volumes
Schedule – this is the conversion of the plans into running sequences for each resource in the factory that is part of delivering the plan. This is where is starts to become hard work as there are many factors that need to be understood to effectively schedule one single production line, let alone many
MAV (Mean absolute value) – not widely understood but a measure of forecast accuracy. Ideally, we measure forecast accuracy at SKU level not overall, and here is where the bun fight starts
- Lets say we have 10 items in our SKU range and we forecast to sell 100 of each, that’s a total of 1000 units.
- The order comes in for a total of 990 units,
- the commercial team are happy they have 99% accuracy
- But at SKU level we get 9 products with 10 ordered and 1 with 900
- this gives you a mean absolute (MAV) of just over 10%
- we have 890 units going in the bin and 9 products that will need making in overtime
Let’s show the planning and scheduling dilemma in a ready meal food manufacturing scenario:
Weekly supermarket sales are cyclical with lower sales at the start of the week building up to busier days at the end of the week. Orders will match that cycle plus working with a 2 day lag to allow product to get through their supply chain.
If you are making a product with 10 day shelf life, the retailer will want 8 of those days in store, which means that the order you get at lunch time today will be leaving the site at 6pm on a lorry.
Let’s work this through: You will have fresh ingredients arriving with 1 day life and frozen ingredients defrosting 2 days in advance of use. The order we get on Thursday has started in production on Tuesday, against a forecast, that is 50% accurate.
That can only mean 1 of 3 things…
- You as the food manufacturer doesn’t make enough and shorts the retailer
- Break your factory financially fulfilling the order at the detriment of efficiency
- You manufacture too many and throw the leftovers in the bin
You will start to see now why the commercial team don’t want to get involved in daily forecasting and the 2 sides spend their lives arguing (hence why you need to bring in outsiders).
Let’s continue with our scenario.
The food manufacturing factory in our example has 200+ SKUs for various retailers made up of a variety of product types and production lines.
We need to agree the scheduling rules…
- What is the preferred line for a product (and a backup line just in case)?
- What is the best sequence for each production line (taking into account allergens, colours, flavours, tray sizes, production tooling equipment)?
- How many people do we need to run the line?
- What machinery? Is it being used elsewhere?
- How fast does the line run?
- How long does it take to change over?
- What materials are needed?
- Are there time constraints on WIP holding time? etc
….Your average production scheduler isn’t going to know this stuff.
We then need to execute the various plans: what happens if one lines runs behind? What if people don’t turn up? What if we have a breakdown?
The factory will need to be rescheduled, and this will happen numerous times on a shift hence why there is usually little desire to schedule to this level. That in turn creates a huge problem in that we have no idea how efficient you should be, or how well you are performing each day.
Still with me?! Let’s go further:
Most food manufacturing operations will have a split between low & high care areas (or prep and finishing). Low care areas include:
- defrosting meat
- weighing ingredients
- making spice batches
- preparing veg
There will also be a kitchen area making sauces; ovens cooking meat; potato cooked and mashed; pasta cooked and chilled.
We then move to the high care area where the production line hosts the final stage to assemble the ingredients into a tray and put a lid on it.
All these areas need aligning in a schedule. What do we usually find:
- There is only a schedule for the high care, leaving the rest to Operations
- Different schedulers looking after each of these areas with very little communication
So where does Chasm find your wins?
These areas are badly run, very inefficient and running overtime and extra hours to fulfil a plan. This is where we find up to 20% improvements in labour.
- First we build the demand planner to convert sales forecast into a plan
- We build a planning tool that takes all the known factors and gets accurate splits of daily volumes to convert into the base data for the downstream operations
- Next we schedule the lines, and we need operational input here to agree lines, run orders, crews, speeds etc.
- Then we get them to execute the plan exactly as it is given to them and measure performance of that delivery to find the standard
- Create a nice feedback loop with schedulers to modify master data as we learn from the process and continually adapt as the environment changes
Doing all this long hand, in Excel or even worse on fag packets and the back of envelope, with poor communication is never easy. It takes time, and knowledge, and effort.
That’s where Chasm come in, with the conflict that planning and scheduling brings up internally, outsiders are the best to fix the planning and scheduling dilemma.