Stochastic Planning – A Shift in Paradigm from S&OP
With stochastic planning and ability to learn patterns of behavior we are at a point to know the exact availability of resources, tools, customer and supplier behavior as they change in different seasons, with events or other social and environmental factors. This knowing enables accurate and adaptable modeling that leads to producing plans that are not just optimized but exact enough for immediate execution without user intervention.
S&OP technology is a rough modeling of the supply chain that does not take into account the variability of resource availability and delivery of raw materials as well as likelihood of disruptions. Everything is based on averages and rough estimates of such parameters. This approach might work for high level and long-term planning but when it comes to having plans that are executable today, next week or even next month, then a lot of manual adjustments are needed. This is not digitalization and nowhere close to autonomous planning. It is simply rough-cut planning!
With stochastic planning and ability to learn patterns of behavior we are at a point to know the exact availability of resources, tools, customer and supplier behavior as they change in different seasons, with events or other social and environmental factors. This knowing enables accurate and adaptable modeling that leads to producing plans that are not just optimized but exact enough for immediate execution without user intervention.
Consider a critical parts supplier that delivers on average in 14 days. However, a closer examination shows that during summer season the delivery is more likely to be between 10-12 days. Planning based on the likelihood of earlier delivery enables more accurate planning and even faster delivery to your customers and of course less inventory. As another example, consider the likelihood of a bottleneck resource availability in the last 5 days before maintenance. If the probability of breakdown increases during the last 5 days, realistic plans can be generated and better delivery dates are given. Furthermore, plans are made in a way that changes the mix of products avoiding heavy use of the bottleneck resource. Lastly, in winter months the transportation leadtimes may show a probability distribution that indicates higher occurrence of longer leadtimes due to explicit weather conditions in certain parts of the country. To this end, a realistic plan is generated to take into account the likelihood of snow storms.
Above demonstrates the value of S&OE and generating realistic models that are true digital twins of the supply chain leading to better plans, higher customer satisfaction and lower cost of operations. For more information on S&OE please click Here.
