Dr. Yossi Sheffi, director of MIT Center for Transportation & Logistics
Artificial intelligence (AI) will have a major impact across the retail enterprise.
Chain Store Age recently spoke with Dr. Yossi Sheffi, a global supply chain expert who is director of the Massachusetts Institute of Technology (MIT) Center for Transportation & Logistics, about how retailers can leverage AI to prevent and mitigate supply chain disruption, among other uses. Sheff is also the Elisha Gray II professor of Engineering Systems atMIT.
Can retailers use predictive AI to help ease supply chain disruption?
Predictive AI would not probably be used to ease disruptions once they happen, beyond allowing sometimes for faster response. In most cases, algorithmic responses may actually not work if they were trained in normal times.
However, AI can be very useful in anticipating supply chain disruptions. The ability to process not only numbers, but texts and images, means that every text and every image around the world can be scanned continuously to highlight items such as such as redundancies, key executives leaving the company, failed merger & acquisition activity, and lengthening terms of payments to suppliers.
All these factors indicate stress, and the supplier may have difficulties that should be checked before it has problems supplying critical parts. Furthermore, unlike financial statements, which are backward-looking, these are real-time alerts.
How will generative AI change transportation and logistics?
The move to autonomous cars and trucks started well before generative AI. Generative AI, in its current form, may simplify and automate administrative document exchange. Imagine scanning items collected in a warehouse to immediately create manifests, bills of lading, certificates of origins, export licenses, packing lists, or countless other forms of supply chain documentation.
What can retailers do to ensure the success of generative AI-based customer service chatbots?
Test, test, and test again. Make sure that managers analyze the interactions and even “listen” in real time to understand how consumers respond to these bots. Many of the bots are going through tedious menus, do not understand what the consumer wants and frustrate everybody involved.
Making the bots easy, making them understand what the consumer wants in terms of explanation rather than menus (buy understand multiple accents and speech patterns.
Beyond chatbots, what retail jobs might be performed by AI in the near future?
Checkout functions are already automated. But some other examples are activities such as navigating the store, creating suggestions for buying complementary products that are sent directly to consumers’ devices while they are in the store, and generating spot promotions based on real-time buying are a few examples.