Updated: Sep 24, 2022
There are a few ways problems are solved quantitatively. Before computers, analytical solutions use mathematics to derive formulas that produce a solution based on a set of variables. Economic Order Quantity (EOQ) is an example of an analytical solution.
When computers became available, more complex problems can be solved using methods like linear programming that optimize outcomes given a set of variables with predefined parameters and their relationships to the outcome. For even more complex scenarios, computers can be used to create simulations to find patterns and probabilities of outcomes that assist decision-making that is difficult to optimize by brute force algorithmic or heuristic methods.
Artificial Intelligence will be able to increase the complexity of problems that can be considered. Not only will it be able to consider a greater number of variables for a specific problem but with the incorporation of big data, it can consider variables that are adaptive to data input rather than fixed by user definition. This technology is powerful because it can improve by learning from data. AI technology is still in its early stages and new applications are still being explored. Here are three areas that AI may be applied to:
Predictive analytics – finding patterns for demand forecasting and then making better forecasts will reduce risk and volatility in the supply chain. This can be applied in distribution and retail businesses to optimize inventory or it can be used in manufacturing businesses to assist maintenance planning of operating assets.
Automation – AI can be incorporated into robotics for safer and more efficient operations. Self-driving trucks may be able to relieve the truck driver shortage that has been persistent in North America for the past few years.
Optimization – many optimization problems in supply chain management are NP-Hard problems like the Traveling Salesman Problem. AI may be able to provide better solutions for these types of problems.
AI can also indirectly support supply chains. For example, AI that is being applied to material science can produce packing and packaging materials that improve shipping and handling. Additionally, better designed products enabled by superior materials can put less strain on the supply chain by increasing durability or by providing a tighter match between supply (what is available) and demand (what is desired).
I haven’t been directly involved in any AI projects to date so my expertise is limited in this area. For some examples of projects underway in Canada and other information regarding AI, you can visit scaleai.ca. Scale AI is a public-private partnership based in Quebec for innovating AI in the supply chain space.