Imagine you drive your usual route to the office every morning. And now in spring, when many people switch from public transport to cars, the traffic on the road gradually increases every morning. This leads to longer queues at traffic lights, slower average speeds and, as a result, late arrivals at work.
You will probably react to this by setting off a little earlier; over time, you will gradually start driving a little earlier still. As the other commuters react in a similar way, after a while you find yourself in the same traffic jam with the same fellow travelers as before; only half an hour earlier. And yet you are still late for work.
It becomes really problematic and impossible to plan when there are lots of police cars, ambulances or fire engines with flashing lights on the route. To the extent that they ‘force’ the right of way, the situation becomes even more unsafe for other drivers. At some point, you will probably give up trying to leave a little earlier and accept being a little late for the office.
What you may be familiar with from road traffic also occurs in a similar form in production control and manifests itself as a typical error cycle in production control: incoming orders increase and more production orders are released. Queues at the various processing stations are increasing and the proportion of orders that are completed on time is decreasing. Production control intervenes and “chases deadlines” by prioritizing particularly important production orders and pushing them to the front of the queues at the production facilities. Deadlines can be saved for a few orders in this way, but the majority of orders tend to be completed even later. However, if my production orders are constantly being completed too late, as the production controller I have no choice but to schedule them earlier. This inevitably makes the queues even longer, the work in progress in production increases, there is even more prioritization and adherence to deadlines decreases even further.
As in our traffic example, production control eventually gives up trying to schedule orders even earlier and hopes to master the situation by chasing deadlines. The error rate in production control is stabilizing at a high level.
In one of the companies we supported, we found throughput times of around four weeks in mechanical processing, although the work content of the individual production orders, including set-up, was barely more than ten to twelve hours (!) at peak times!
A comparison of the average number of production hours required per week over the past few months with the number of production hours relieved per week showed that the average rate of additions and disposals was the same.
This is a typical effect of the production control error loop: despite the long throughput times and work in progress in production, on average no more work volume per unit of time is put into production than flows out of production.
How can you avoid the production control error loop? It would only be theoretically possible to stop putting production orders into production. On the other hand, it makes much more sense to adapt production capacities better – and above all promptly – to demand. In practice, however, the response to increased capacity requirements is usually far too hesitant.
However, if you respond to increased capacity requirements at short notice by increasing available capacity, you avoid the production control error loop and keep the rate of inflows into and outflows from your production in balance. You should then reduce the circulating stock to the level actually required. To do this, you need to temporarily expand production capacity through various measures: Outsourcing orders, overtime, weekend work and ghost shifts are some possible levers. In the company I mentioned, it took us around six months to reduce lead times from four weeks to four days in this way.
You can precisely measure the right production work in progress for your company for strategic and competitive reasons, but this is a different matter. With the above approach, however, you can already avoid congestion in production. Have a good trip!
