sep

Why are so many people afraid of getting on an airplane but will never hesitate to hit the accelerator of their car? Do you also prefer to avoid the idea of confronting a shark after seeing some of the horrible scenes in the movie Jaws? But why do we then classify landscapes with cows as idyllic? Have you ever heard those never-ending success stories about strategies to beat the casino? But why do casinos then still exist?

Uncertainty is everywhere
The above questions seem somewhat arbitrarily formulated. But they all have one thing in common: humans are intuitively not the best when it comes to probability decision making and statistics. Commercial flying is statistically safer than driving [1]. On a yearly basis, more people are killed by cows than by sharks [2]. If you have ever outsmarted the casino: you should either shut your mouth or you are part of Ocean’s Eleven!

Uncertainty plays a key role in our daily lives. It can be accurately quantified in a more stylized setting like a casino. In theory, you can always roll 6. On the other hand, in most cases, you only have some form of approximation or feeling based on historic data and experiences. Moreover, the source of uncertainty is often overlooked. It is not all about whether you can do the math, but also about understanding why not everything is 100% predictable and knowing if you have any control over the process.

Spare parts and uncertainty
We at Gordian let us also guide by, among others, quantifiable approximations and statistics for spare parts inventory management. Supply chains are generally subject to many forms of uncertainty (not extensive):

• Demand/Breakdown uncertainty
• Supply uncertainty
• Production/Repair uncertainty
• Resource uncertainty
• Yield uncertainty
• Shelf life uncertainty

It is important to distinguish processes on which you can exercise influence to some extent. Of course, one can somehow identify why and when machine parts fail at a certain point in time. Similarly, one can extensively try identifying why a customer buys this typical product at a specific point in time. However, in supply chain management, there is usually limited control on the demand side. In contrast, internal production, supply and repair processes are steered and controlled by you. Too much variability would complicate any form of planning activity. As a result, companies steer for stable internal lead times (e.g. Toyota KANBAN-systems).

The big mistake
Even though internal process lead times are generally quite stable, the time between work order request and receipt can be rather variable. This is often due to the unavailability of materials, caused by a stockout at a preceding supply chain stage (higher echelon level). The source of uncertainty is not the process itself, but rather unforeseen high demand at the end of the chain. Hence, doing the wrong math may overestimate the amount of uncertainty in the supply chain. This principle is demonstrated in the picture below:

Remember
Do not be too fixated on statistics; local information is valuable. When you start understanding the source of uncertainty, statistics are a great building block for spare parts optimization. Gordian aims for balancing machine availability, inventory capital and operational costs by using both quantitative and qualitative information. We do not get distracted by short-term successes or bad luck. Hence, think twice before you start challenging the casino: the odds will probably turn against you in the long term.

[1] https://www.lesserlawfirm.com/is-flying-safer-than-driving/
[2] https://abcnews.go.com/Health/shark-versus-cow-deadlier/story?id=24931705