The evolution of  Supply Chains will be a step-wise process. What those steps will be will, and should be, is a unique narrative for every organization. But eventually, we will have supply chains where algorithms will develop models to take certain decisions. And these models will flux as the underlying data changes. The types of decisions will be based on criticality level- more critical ones will still need human review. Logistics is an area ripe for this algorithm driven decision making. And many of these decision models developed by algorithms will be classification models.

Logistics Operations- Decisions Galore

Many of the day-to-day aspects of logistics operations are instances where someone needs to make a decision. This decision eventually translates into a “Go” or “No Go” next step. Note that when I say logistics operations, I am excluding strategic and operational logistics planning aspects (for this context)-which are currently primarily already driven by optimization algorithms. This article focuses on tactical operational decisions.

Classification algorithms (a brief tailored explanation)

I will explain what classification algorithms in the world of machine learning are from a supply chain and logistics perspective so that you don’t have to go through all the jargon that is often thrown (sometimes on purpose) at you when the discussion pivots to machine learning.

When you make certain decisions in logistics operations, you have specific options, and you choose one of them. An example can be that you have an outbound load staged at the door with all the products but one pallet that is Inbound but still hours away.

As an operations manager, you need to make a call- wait till the product arrives or dispatch? The decision classes, every hour, are: dispatch or do not dispatch.

To make that decision, you take into account specific data points- how far away is the Inbound, what products are on the load, for which customers, what are the chances of arrival delay- and based on that, you assign an overall “chance” – if the “chances” of the outbound load getting delayed exceed a specific notion in your mind, you decide to dispatch, even though the Inbound load has not arrived.

This is what a classification algorithm in machine learning does. It calculates the “probability,” based on that; it classifies the outcome into a class- in this case, dispatch or do not dispatch. All the data points you took into account will need to be fed to the algorithm through automated feeds so that it can then suggest and make a decision based on the criteria assigned.

Explaining application opportunities with examples

So the logistics operations part is where algorithms must pitch in to make decisions. In my mind below are some examples:

Example 1:

An Inbound product is currently in transit, and an outbound load is waiting for this product; a decision needs to be made on whether it should remain after a specific time.

Based on the real-time location of the Inbound load, expected arrival and loading time, etc., a classification algorithm like Naive Bayes can assign the probability of arrival delays for the outbound load based on the departure time (multiple departure time options). And if the logistics managers define a cut-off threshold fed to the algorithm- the algorithm can then propose whether the outbound load should wait for the Inbound.

Example 2:

Will an outbound load get delayed?

This is a binary Yes or No decision. Logistics regression classification methods can be used to predict the “class,” which is “delayed” or “not delayed,” based on specific attributes like lanes, product, carrier, driver (in the case of the internal fleet), vehicle type,  etc.

Example 3:

Will this product get damaged if packed or stored in a certain way?

This, again, is a binary Yes or No decision. Logistics regression classification methods can be used to predict the “class,” which is “damaged” or “OK,” based on specific attributes like the product, Quantity, Primary packaging, Secondary packaging, Transit time, Transit lane, vehicle type, etc.

Conclusion

The list is long, and the opportunities are endless. Every Logistics operations scenario where someone needs to make a “Yes” or “No” decision can be assigned to a classification algorithm. The key is carefully defining and categorizing processes that should be transitioned to algorithms. These assignments must be made years before you get to the actual “smart supply chain” level so that humans in the processes are comfortable trusting algorithms making certain banal decisions in their daily operating environment.