Production scheduling

One of the underappreciated functions in companies is scheduling and production control.
We invest heavily in order processing and engineering software, and we install new, high-tech, Industry 4.0-compliant equipment on the shop floor. Scheduling and production control is the interface between those pillars of technology, and we often fall short on applying the same amount of technology, ingenuity and competence.
Also, with the move towards more integrated manufacturing equipment (i.e. automated panel retrieval systems combined with saws and nesting routers) the importance of competent planning and “frozen” schedules increases.
There is no “one fits all” solution for scheduling and production control. Every sub-category in our industry is vastly different in their requirements. A kitchen cabinet plant is significantly different than, for example, an architectural millwork company. A moulding mill has different needs from a furniture manufacturing company.
When looking deeper into each of the sub-categories you realize that, for example, a furniture manufacturing plant, distributing to a dealer network, versus predominantly selling on-line must align the scheduling system to the manufacturing and business model.
This article will focus on challenges and solutions for the mid-size kitchen cabinet companies, but in part the solutions will apply to any other manufacturing plant.

Overall objectives
Order entry and engineering documents and converts customer requirements into the technical format required for production. Scheduling or production control needs to process the technical customer information and feed production with the following objectives:
• Fulfil or exceed the customer expectation (i.e. on-time, complete, correct)
• Optimize the use of production resources (i.e. feed the plant with maximum work)
• Minimize the manufacturing cost (i.e. minimize over-time, provide level loading…)

Basic prerequisites
To give scheduling a chance to “drive” the plant, we need to provide the technical information (work orders) on-time. For example, if the saws or nesting machines are repeatedly running out of work and need to wait for the next order, scheduling can only issue the available orders and cannot optimize for production.
A second prerequisite is to provide highly accurate information. If the engineering and order information has errors (false information, missing information…) and the errors are discovered while the orders are being processed, it will interrupt the flow. And every time an order stops on the shop floor, productivity is killed. A few errors can be covered by the agility of the production team, however, if it surpasses a certain threshold it will bring production, and definitely productivity, to a grinding halt. It also moves control from scheduling to production. If too many work orders are incorrect, the production supervisors will work on the orders with enough information to proceed. This process, I call “riding the chaos” is not sustainable in the long-term.
So, in order to develop a properly scheduled plant we need to stabilize and fix the data flow to the plant first.

Optimizing production
A complex plant, such as a kitchen manufacturing plant, can be metaphorically compared to an orchestra. It will not work if every instrument is completing their part as fast as they can play it! But many companies do this. Orders are issued to the shop floor as fast as they can. Machines, i.e. saws and nesting routers produce as fast as they can and so on. The work in process (WIP) flows slowly and will back up at bottleneck operations. This leads to the shop floor being stuffed with material until all transport utilities (roller conveyors, carts, pallets, skids…) are occupied. I have seen in extreme cases, re-piling of orders to gain empty carts, and workers chaining empty cart to their workplace. Everybody calls for a bigger plant and more space, more carts and more machines! Again, the productivity suffers.

Know your bottleneck

You need to know what your bottleneck is. As you investigate your bottleneck(s) you will see there are many “false bottlenecks.” They are false because they can be opened relatively easily by shifting resources. Once you have identified the true bottleneck you will know how much the factory can produce. You only can produce as much the bottleneck allows.

Required TAKT time

What is the customer requirement over the next short and medium terms? How many cabinets do we have to produce to fulfil the requirement (cabinets per day or per operating hour…)?
The comparison of bottleneck capacity to TAKT time shows if the plant has the capacity to produce the customer requirements. If the requirement exceeds the TAKT 
time capacity and orders cannot be moved out (later delivery), management needs to find other means to increase capacity (i.e. overtime, extra shifts, outsourcing…)
Issue manageable work
Now that we have balanced the maximum capacity, we issue orders in manageable daily batches. Manageable also means that they will process through the bottleneck(s). In a typical kitchen cabinet manufacturing plant, we might have multiple upper limits:
• Overall number of cabinets per day
• Number of cabinets with wood or painted doors (finishing capacity)
• Number of cabinets with special finish doors (high gloss, antiquing, special effects…)
• Amount of work for custom shop (Specialty hoods, etc.…)
If we constantly enter larger quantities per day than those established upper limits, orders will pile up on the shop floor and the on-time performance will be in jeopardy.
For example, the scheduler must not issue more orders for wood and painted doors as the upper limit. However, he/she can shift some orders with laminated fronts (or out-sourced fronts) to top up the daily upper limit. We refer to this daily quantity as the “daily bucket.”

Complete your 
daily bucket

As the work orders are issued to production, each production group focuses their work only on that order. They should not start the next day’s order if it is not on their schedule. The textbook process answer is that once their daily objective is completed, they should stop working and, if possible, shift their resources to a department behind on their daily workload. They could also work on improvement projects (as we teach in our Bluewater Wood Alliance Lean Training; Brad Cairns/Sepp Gmeiner). Some other companies allow their workers to clock-out and go home early.
In our sample kitchen cabinet plant, the first production department of the process value chain (saw and nesting routers) has completed yesterday’s workload on-time, at the end of yesterdays shift. The new bucket for todays work can start again at the beginning of the day and should complete at the end of the day. The next operation steps, like edgebanding, drilling etc. can start almost simultaneously with only a small time-offset. Within a few shifts or days, the entire daily bucket moves through the factory. The length of the factory through-put time for the daily bucket only depends on how much time you allow for the different offsets. The fewer of the above-mentioned errors and interruptions you can insure, the more overlap of operations (less time-off set) you need to allow.
The impact of such controlled production is a radically shorter factory through-put time. Also, the space requirement is reduced because you have much less WIP in the factory. This will also calm the production process, production snags become immediately visible and productivity increases substantially. Achieving such a state of control has high rewards, resulting in on-time performance, capacity increase, productivity increases, quality improvement and higher profitability.
To achieve such planning, most of the improvements to the current system need to be done in the upstream processes.
 
Fill the buckets
The ability to give the production planner/scheduler elbow room to plan requires that, at the point of finalizing tomorrows daily bucket, the planner has more orders available to schedule than needed. This requires that we actively drive process completion in the upstream processes. Order entry, field measurements, colour selection etc. need to be streamlined as much as production processes need to be improved. With improving on both ends we can increase the pool of orders ready for scheduling.

Outlook
Developing and implementing your own scheduling system is a long process. There are suitable industry software solutions available, but having the right software is only a part of the solution. There are no “plug-and-play” solutions available. The successful implementation and the required culture change are the more difficult parts.  Competent employees for all related departments need work together to design and build their scheduling system.
And when they have successfully implemented their system and are running like a well-oiled machine, I would challenge them to switch from daily buckets to half-day buckets. This shifting of gears would open new improvement potentials. But get the one-day bucket implemented first.