Workload Control is a centralized WIP (Work-In-Process) management methodology for High-Mix Low-Volume production environments. It provides a consistent framework for releasing orders by taking into account the WIP for each team or workcenter.
In manufacturing, it is important not to release too many orders at the same time, as you may build up an excessive amount of WIP on the shopfloor. On the other hand, releasing too few orders may lead to workstations running idle. Hence, one needs an approach to release the right amount of orders and balance throughput and WIP. This is what Workload Control is all about. Workload Control helps you avoid the trap where you release orders just to make sure everybody is busy. When implemented effectively, Workload Control will maximize throughput & output while keeping WIP under control.
The figure below illustrates the underlying idea. Beyond a certain threshold, releasing more orders does not increase output any further. Hence there is an ‘optimal WIP’ that maximizes output while keeping the production lead time as low as possible (Little’s law states that lead time equals WIP over throughput).
WIP vs Output (source: PLANVAR).
Workload Control is a concept developed by Land, Thürer & Stevenson in 2009. Numerous academic papers have been written on the topic, introducing variations of the same concept and presenting performance results, mostly based on simulations. In the main, here is how Workload Control works:
The left-hand side figure represents the (normalized) workload for each team, based on orders that have already been released. The target maximum workload is called the workload norm. Now assume we consider releasing an additional order, of which the workload is shown in the middle. This new order requires capacity from teams A, C and D but does not pass via team B. The right-hand side shows the impact if we would release this order: the workload of both team A and C would exceed the workload norm, so we may consider releasing other orders first, effectively limiting the workload for each team.
In sum, for setting up Workload Control you need the following components:
- Workload norm — a target amount of work for each team, e.g. 2 days of work. If the team is bigger, or the workcenter comprises more machines, the workload norm will be higher for that team or workcenter.
- Current workload — for each team, the current workload is the cumulative normalized workload of all orders that have been released and still have to be processed by that team. It is based not only on the orders currently in the buffer at that workstation, but on all released orders still on their way to it. When a team finishes the work for an order, that order’s workload is subtracted from the team’s workload. Normalized means a weighting factor is applied: the further the team is in the sequence of operations, the lower the weighting factor.
- Release algorithm — when deciding which orders to release to the shopfloor, the Workload Control algorithm checks whether releasing a specific order will push the workload of one or more teams beyond the workload norm. The idea is to release the orders that keep each team’s workload under or around the norm, and so balance the load across teams.
Workload Control can be implemented in an analogue way, but a digital alternative is more straightforward.
In 24Flow, Workload Control is used in conjunction with Order Pool management. The Order Pool is a ranked list of orders to be released, from which planners or team leaders select orders for the shopfloor. Typical criteria taken into account are order priority, planned start date, order due date and availability of components.
Workload Control nicely complements the Order Pool: it adds the WIP of each team as an additional criterion for deciding on order release. Have a look at the video below to see Workload Control in action in 24Flow.
See how Workload Control keeps WIP under control on your shopfloor.