So called autonomous supply chain planning is all about improving planner productivity. Will the day come when a system can perform all your planning functions and respond to events without human intervention, just the way you want it? Yes, but not soon. Automation simply helps to improve productivity, and in most cases relieves humans from laborious and repetitive tasks that require either little thought or involve some complex mathematical calculation with thousands or millions of variables, so called optimization. Operations Research (OR) techniques help in the latter category. AI/ML techniques, including Gen AI, can be of value in the former. Planners spend a lot of time reading and responding to emails, many of which are mundane and time consuming. They also spend a lot of time trying to reconfigure orders and respond to changes that come from customers or disruptions, internal or external. They do so by examining scenarios. Only a handful from thousands of possibilities. Not a good use of their time and expertise.
Systems can and should indeed be capable of performing the scenarios automatically to find the desired solutions and present them to the planners. Systems can also learn from planners as to what is important and what is not. Thus, guiding the planners to spend their time on what matters or what the system does not know what needs to be done. To this end, systems act as a faithful assistant to the planners helping them to focus on more immediate and higher priority tasks. When a message arrives to a planner that says the shipment is late, they have to spend much of their time to figure out if this is relevant or has any impact on the plan. Systems are perfectly capable of doing this in real-time and making recommendations to the planners one way or the other.
More importantly with the use of recent Large Language Models (LLM), it is possible to interpret the nature of external messages to automatically distinguish the extent of any relevant disruptions. Hundreds of events are happening around the world that may or may not be relevant or impactful. LLM can interpret the message to understand what the issue is given its many forms of natural language. Thus, the system can tell the difference between an oil spill in a supplier site or oil shortage for the delivery trucks. It is then up to the planning engine to figure out the relevance and impact thereof.
Using the systems to make the current planning processes faster may improve planner productivity a little. However, understanding the power of current technology to establish a whole new process should be the focus. The possibilities are endless and the gain in productivity is a quantum leap not just for the planners but also your operations. Such would be your competitive edge.