Defining the AI Supply Chain Imperative

Way back in the year 2000 I had already identified the potential for Artificial Intelligence to have a major role in supply chain management and at that time even began tracking the topic on my About.com Logistics/Supply Chain site where AI was the first topic due to the alphabet!

Little did I know then that it would actually take 24 more years for this technology to finally mature and be ready for primetime in the mainstream business world, but like a fine wine it seems some things just take time!

And most importantly, regardless of how long AI has taken to get here, the inverse has now become true, and the speed at which the need for and applications for AI in the Supply Chain are moving is nothing short of jaw dropping.

The end result of this is what I'm now coining as simply:

'The AI Supply Chain Imperative.'

So just how do we define "imperative" in this context?

Definition: imperative(n) [im-per-uh-tiv]?noun: a command, rule,?duty, etc., that is very important or necessary.?Serious edict issued with or from position of authority or advanced knowledge.

The imperative in this case is that in the immediate future utilization of AI in your supply chain won’t just be a “nice to have” but will become an absolute requirement to ensure end to end monitoring, optimized performance and costs, as well as providing ongoing risk assessment, identification and the orderly mitigation of any risks who’s impacts can’t be avoided.?

?In order to be more specific about these Artificial Intelligence application areas I’ve recently conceptualized and coined this as The MORAM AI Supply Chain Cycle.

The MORAM AI Cycle for supply chain management is depicted below:

?

?

Monitoring

Any supply chain output sensor or system driven can be captured, tracked and actioned as appropriate. Examples of these include IoT of all types, Vehicle ELDs, Temperature Sensors, AI Cameras for OTIF/Claims tracking, Shock or other devices with data of value. Additionally, any directly correlated data such as service level confirmation, carbon monitoring or other supply chain output can also be captured and recorded for reporting purposes.?

In addition to the sensor data sources above, there are literally 1,000’s of other supply chain relevant data points within the below general categories which can be monitored and input into AI Supply Chain models including:

?

The above represents current potential input sources and there will be additional discovered going forward and these will evolve over time.

?

Optimization

As one can imagine, with all of the data being monitored, collected and stored as described above, the opportunities and different types of possible optimization are numerous.

Optimization is of course based on bias of the optimizer to a given situation / scenario. For example, organizations can optimize for their supply chain cost even when this might result in higher costs for other supply chain participants and even the end to end supply chain as a whole.

The same applies when factors such as Carbon are optimized as this may lead to impacts on both cost, service and schedules.

The below chart highlights some of the potential Optimization types which can be embedded and integrated.

?

As you can see there are multiple possibilities here on how AI Optimization can be applied to the supply chain which will need to be driven through the specific goal setting of those managing any given supply chain which is to be AI optimized.

Additionally, all of this raises applications to wider supply chain questions around items such as location sourcing of products, manufacturing locations and batch sizes, on hand inventory quantities and stocking /deployment locations which are beyond the scope of current article.

Risk Assessment

Within every supply chain there are many risks and every action / movement within the supply chain can be assessed with a numerical measurement of risk from “100% going to happen” down to “0% will not happen” and every point in between these.

Actual Risk Assessment of these supply chain activities can be calculated from the 1000’s of supply chain data points reviewed within the Monitoring section above with the end result being a Risk Assessment score.

Each time a Risk Assessment is calculated represents an opportunity to potentially take action to improve this score. By doing so reduces the risk of 0% outcome or failure of the given supply chain activity.

Such actions reduce the risk of supply chain failure and is the first of two types of Mitigation to be discussed next, the first being Remedial to try and prevent the supply chain activity failure from happening and the second is Recovery which occurs after the supply chain activity failure has happened. ?

?

Mitigation

Based on the information identified above, for every supply chain risk identified, specific mitigation rules and methods can easily be defined and automated through Artificial Intelligence to quickly address and resolve any of the identified supply chain risks or issues based on business rules / logic specified.

Two types of mitigation can be performed, firstly Remedial Mitigation to take actions to reduce risk of an impending supply chain failure. And secondly Recovery Mitigation to take the most orderly path to resolve supply chain failures after they occur.

As it’s only a matter of time before Artificial Intelligence in supply chain becomes the norm instead of the exception, the time is now to determine how you and your team are going to best leverage the use of AI in your supply chain.

Adopting comprehensive AI models for your supply chain based on frameworks like The MORAM Cycle explained above are a great initial step in ensuring all are ready to conquer the challenge now upon us, The AI Supply Chain Imperative! ??

?

Jeff Ashcroft?????????????????????????????????????????? ?????????

416.990.6433

[email protected]

Ready to discuss the application of AI to your supply chain? Just reach out!

If you enjoyed this post click here to see All my Articles, Follow or Connect with me on LinkedIn! Cheers!