Planning with uncertainty
The current Covid pandemic requires us to make decisions all the time. Underlying these decisions are facts and statistics about how the pandemic develops, and expectations about future events, like virus mutations, vaccine effectiveness and the effect of restrictive measures. Similar to many production and transportation environments, these future events are surrounded by uncertainty. This is similar to a supply chain planning context. The strength of a human planner is that (s)he is able to deal with uncertainty, whereas an automated planning system is not. However, not all human planners are equal, and dealing with uncertainty is a skill that can be acquired.
In 2004, Ken McKay and I published a book that could serve as a ‘survival guide for planners and schedulers’. In the book, we described ways for planners to deal with unexpected events, and the ‘deadly sins’ a planner could commit in dealing with them. A planner can be seen as a manager, as (s)he makes decisions about some operational process. However, planners typically make many more decisions on a daily basis than managers, and therefore are probably more skilled in making decisions in an uncertain environment.
One of the ‘deadly sins’ that a planner can commit, is to not have a plan B – an alternative plan when the principal plan fails. For example, a planners needs to make sure that some important customer order is shipped, and (s)he needs a specific production line for that. However, (s)he should also plan for the situation where the line will fail to operate, or produce the wrong quality. Is it perhaps possible to use an alternative line, can we rework some inventory, can we negotiate a later due date?
What concerns me immensely regarding the decision making around the Covid pandemic, is that the principal plan – we apply restrictions to society until the vaccine ‘saves’ us – does not have a plan B. A good planner would consider the things that can go wrong and how to mitigate them: suppose the vaccine does not work in practice as promised? The figures published by the manufacturers look great of course, but they have not been proven in a real life situation yet. Or, suppose we get a virus mutation that the vaccine will not fix? Not having a plan B, is asking for problems. The vaccine ‘should work’, but in my work with planners, the words ‘should work’ make me cautious.
Without a plan B, this would mean that our healthcare system might get overloaded (‘code black’), or that civilians that are not strongly affected by the virus will no longer accept the restrictions, or – and I find that more likely – both. Suppose that governments try to impose restrictions that civilians do not want to accept, but civilians feel they cannot go into a discussion about them as they will be facing experts with better wording abilities and strong moral arguments. It is like a key user debating the usefulness of a planning system with a consultant who has a PhD in mathematics. The PhD will win the discussion but loose the case. Theories of organizational change will predict that the restrictions will be ignored, or worse, sabotaged. For example, when a new information system is imposed upon a group of professionals, and they do not understand the benefits of this (even when explained by brilliant consultants), or the benefits will mainly be for others while they carry the burden, they will make sure your project will fail – by not cooperating in the project nor using the system when it is live.
So how would a plan B look like? With the risk of threading into a domain where I do not have all the required expertise, the main measure might be to seriously extend healthcare capacity – specifically for Covid patients. Now the main blocking factor here seems to be the availability of personnel. I believe it will help here to see a hospital in the context of a supply chain production unit – one that can deliver many different treatments. And this unit is now overwhelmed by patients that all have the same disease. From supply chain studies, we know that it is not efficient to have one unit deliver both – a large variety of different treatments, and a large quantity of one treatment. That would be like a car factory, delivering cars, bicycles and ships at the same time.
In such cases, it is better to take out the treatment that is needed in large quantities and to set up a unit specifically for that – not a job shop, but a flow line. A rule of thumb tells us that to produce large quantities of the same thing, mostly routine tasks are needed, with a minority of the tasks requiring attention from an expert. And for the routine tasks, employees can be trained more easily. In other words, less experts are needed per patient in such a setup. Such supply chains turn out to be quite scalable, as you only train employees the skill they need for the routine subtask they are carrying out.
There are probably many issues to overcome, to implement something like this. Some will say it cannot be done and it will harm healthcare quality. However, this could also be said of the vaccine, a year ago – that it could not be developed safely inside a year. When pressure mounts, things that are officially impossible, turn out to be possible. And perhaps the job shop/flow line is not the best plan B to be developed, although I do not see too many other options. Feel free to enlighten me.
Try it out!
When a planner thinks of a plan B, when the principal plan fails, (s)he will make sure to try out this option, perhaps on a small scale, to make sure it works. For example, when an alternative machine is officially not suitable to produce a certain item, but it can be modified to do so, a good planner will assign some small quantity of that part to that machine beforehand, to try it out. Such efforts are currently not done and not even considered, which concerns me greatly. The only measure currently used – apart from the vaccine – is more and tougher restrictions – as if these can be applied without limit. When civilian populations agree to have their life restricted, that is not an issue. It is a bearable situation to many, as we all expect the vaccine to release us. But again, when the vaccine fails, the support for such measures will drop drastically. This means that politicians cannot continue to only push the restriction buttons. A good planner would already start to find out, where the trade-off lies – how much more can we restrict, for how long, how much more capacity can we create, will it work in practice?
The vaccine arrived in record time, and I sincerely hope it will be the measure that ends the Covid crisis. However, when the vaccine turns out to be not the miracle that we expect, we might have no other option than to resort to a plan B, which might mean, increasing healthcare capacity. And when we do not have a plan B, we will have to keep the restrictions on society on for an indefinite amount of time, or until the virus decides, that it has been enough.