AI use cases in supply chain

From Raw Material to Finished Product

According to a 2017 survey of 835 companies, Artificial Intelligence (AI) has penetrated deeply into our everyday business activities. That is not much of a surprise for most of us since the subject continues to hold sway in reporting and gather interest from almost every news outlet.

It tells us, for example, that 44% of companies now use AI to detect intrusions and security breaches. Another 41% are using it to resolve our users’ IT problems with, among other things, amazingly human-like ChatBots. These are the sorts of things that people expect to see when considering AI implementation.

Behind the Scenes

Less obvious are its uses in streamlining our Supply Chain (SC), but they are certainly no less valuable. When even modest-sized businesses are expending almost 60 hours per week manually managing the paper trail for orders, another 40 tracking invoice anomalies (such as discrepancies and errors), plus another 24 hours just to respond to supplier inquiries, it quickly adds up to more than 6,400 hours per year.

Considering that we each typically work about 2,000 hours per year, and we seldom achieve even 50% efficiency (accomplishing something important one minute out of every two), this translates (roughly) to the output of at least seven full-time employees (in the most efficient organizations) dealing with this on a full-time basis.

AI Automates the Mundane

There is a way to escape this time-stealing vampire that sucks the productivity out of your business while causing employee burn-out… Modern AI can eliminate virtually all of the repetitive, mind-numbing work, yet as recently as 2014, AI wasn’t even included in the annual SC analysis survey!

A good 34% of companies have already turned to AI to reduce Production Management workload by automating it, and this had freed up those employees to use their creativity in new but related areas. Since 2016 we’ve overcome much of our internal resistance to SC-AI, and now the early adopters are gaining palpable advantages.

AIs can track any production, supply, or shipping anomalies instantly, with 100% dedication to dozens of tasks concurrently, because of the sheer speed of our current computers. Once that trail becomes entirely electronic instead of paper-borne, nothing can track it faster than the AI-driven machine itself.

Provided with the time to develop deeper insights, about 20% of companies have begun using AI to predict what their customers are going to order. This reveals what supplies will be required and orders can be placed earlier, assuring priority.

The AI then follows up with targeted advertising to help steer clients in the predicted direction even faster. Customers react by thinking “How did they know I was going to need this? This company is amazing!”

Knowing what Customers are Thinking

Customer responsiveness is tracked by the AI which orders materials on a JiT (Just in Time) basis. That means materials arrive precisely as the machines need it, so it doesn’t need to be warehoused and stored until demand catches up with supply. Instead, your demand matches your supply so your warehouse can remain reasonably empty, or even be eliminated, saving you the cost of leasing property for storage.

Clearly, this is an advantage, since capital is not tied up in materials. It remains available (liquid) for other investments or development opportunities, which means that your company is more responsive and agile than your competitors.


These wonders of the modern age are not just useful for dealing with customers when the questions are repetitive and predictable. The same ChatBots can be used to interact with suppliers and can autonomously requisition materials, confirm shipping orders, issue governance & compliance requirements, receive invoices, and track late shipments.

AIs are perfect for whatever is repetitive and boring because they learn by repetition, and they never get bored.

In their ChatBot form, AIs can even confer directly with floor-production staff when there is a materials shortfall. It can reschedule the production runs in the best way to minimize time and labor, as well as to conserve materials.


Those who fear the science fictional SKYNET or HAL 9000 versions of AI don’t need to concern themselves with our sophisticated and modern (and non-fictional) implementations of AI. We may not have the Three Laws of Robotics, as defined by science fiction legend Isaac Asimov, but we do have brilliant programmers who have been exposed to the cautionary tales of 2001: A Space Odyssey and, of course, Terminator. Even if it is fantasy, let them run-scared of the possibility and program so that our AIs always use a human reference frame to accomplish their tasks.
Smart Decisions

AIs, once properly programmed, are sensitive to the timeliness of shipments, and whether it is better for the customer to receive a partial shipment, or delay production until all the materials are available for a single run. Human decisions are required for initial training, but then the AI can extrapolate from a history of decisions for a particular customer, or a particular type of product, to find the right answer.

We Deliver

An important segment of our Supply Chain is not merely acquiring, having materials on hand, or turning them into saleable goods, but getting them to the customer in an expeditious manner. AIs, connected directly to our fleets, can re-route materials on-the-fly.

It might be weather, traffic, or disaster, but awareness of where our goods are, and which of our customers has the most urgent need, means we can redesignate something headed for Maine 3-days-early, to Delaware, so it arrives JiT, without the customer ever knowing there was a problem.

Other Uses

Natural Language Processing (or “Programming,” depending on the application) allows you to understand languages you don’t speak. Google now supports more than 100 language translations, which as far as earthly languages are concerned, represents a literal Universal Translator.

AIs can use these algorithms to glean new insights from what was previously incomprehensible data. It will be the end of a large block of text being translated as “Deliver Friday.”

Would you like to know what your customer really said, often with contextual and cultural meaning added in? Philosophers have argued that once we fully understand other people, war will disappear… Whether true or not, better understanding means better business relationships, and that means more profit.

Costly Implementation

Of course, some things will be quite inexpensive (like translation) because they run on existing systems, but many items in Supply Chain AI will have a bigger price tag attached. Monitoring and communicating with your fleet, for example, will require hardware for each vehicle and data connectivity.

It may seem to be an intimidating expense, but you should know that according to McKinsey Research, early adopters of logistical AI Systems showed a 5% increase in profitability whereas those who didn’t update were losing profitability for the same period. More profit is a terrific incentive, but leading your competition is even more important because it will increase your market share as they lag behind and struggle.

WATSON Supply Chain AI

One particularly clever version of what we’ve been discussing is the version by IBM™ running on their WATSON AI platform. The major selling point is the synchronicity of all your data. Everything falls into place.

The system interprets worldwide weather patterns to see if something you need will be grown in sufficient quantity; if drought or flood will compromise your conventional suppliers; what alternatives are available if there is a major transportation disruption; and much more, of course.

All of your data, along with publically available information, is analyzed and interpreted to provide an intelligent and highly predictive supply chain story. You’ll know the source of potential disruptions long before anything happens, so you can have strategies ready-to-go if something does go wrong…

The Takeaway

The difficulty we are currently experiencing in supply chain management is the sheer volume of data. It is like a dam that has burst, but in this case, information is flowing fast enough to wash away anything that stands before it.

You certainly don’t want to be the ridiculous tiny figure standing in the middle of a street with a bucket yelling: “Don’t worry…I’ll stop it!”
We have moved far beyond the point where our older systems were still capable of dealing with the information we generated. Supply Chain Logistics is now so thoroughly intertwined with all the details of Sales, Design, Marketing, and Supplier Relationship Management (SRM), as well as geography, culture, and even custom-designed, unique one-off products, that it is completely beyond the capabilities of humans to make fully informed decisions.

You now need AI just to stay competitive, let alone to excel. If you’re late-to-the-party, we would be more than happy to advise how to get up to speed. Please feel free to start a conversation with us here. Our plug and play platform is the ideal tool for end-to-end implementation of AI in your business. We’ll be there for the whole journey, not just a quick sale!