Which Supply Chain Planning Processes Should We Automate?
Now that more companies are embracing supply chain planning productivity and automation, many firms are asking which processes they should automate. Let’s first address that question and then look at where this trend is heading in the future.
Gartner says that the attitude towards supply chain planning productivity and automation has changed dramatically. There is a focus on many companies at either 1) significant (e.g., 30-40%) increases in planner productivity or 2) growing revenue without adding headcount. Large companies with slower growth are under pressure to find the next big bucket of cost savings, so they are often looking to reduce headcount in areas that can be automated. Smaller and faster growing companies are looking to avoid increasing headcount to service their growing business.
Both are looking to automate, but where? Gartner suggests System of Record (SOR) activities such as short-term forecasting where fast-moving daily decisions require lots of data which are beyond human capacity. These are areas, they say, that can be run quite autonomously, sometimes known as“low touch planning” or “algorithmic planning.”
Conscious supply chain planning at the top. Subconscious planning on the bottom.
Some processes have the potential and need to be automatically optimized, while others need to be presented to the planner so they can understand the business trade-offs and make the final decision. Gartner suggests that a good way to look at the question of which processes to automate is to frame it in terms of human brain activity. Subconscious supply chain planning they say can be largely on autopilot to “respond effectively and optimally to relevant environmental signals at lightning speed.” Conscious planning, involving planners supported by scenarios and collaboration, focuses on complex and chaotic situations where no best or good practices exist.
So the best approach is generally not to focus on automating the more strategic activities (Configure and Optimize in the Gartner CORE model) but rather the more execution-oriented planning layers in that model (Respond and Execute). In the top layers there is more conscious planning, where algorithms help facilitate decision making and collaboration across relevant stakeholders. In the bottom layers closer to Execute there is more subconscious planning, where adaptive models automatically apply best practice resolutions as relevant events are detected.
In the top layers, the algorithms should be more focused on resiliency, Gartner says, by creating risk-adjusted plans and scenarios. The bottom layer algorithms closer to execution should be focused more on accuracy, by creating more accurate short term responses.
Looking Ahead: The IoT encourages subconscious planning and vice versa
Gartner says that as we see more transactions from the Internet of Things (IoT), they expect that more planning will move from conscious planning to automated subconscious planning because “the traditional planning approach will not work effectively because of a dramatic rise in events per second from the end-to-end ecosystems that potentially impact supply chains.” With the advent of big data, IoT and smart machines,” Gartner says “organizations need to develop the ‘subconscious’ side of supply chain planning to cope with and make sense of the proliferation of data and events.” (Digital Business Requires Algorithmic Supply Chain Planning, Tim Payne, September 2016)
As the use of subconscious planning grows, Gartner expects it will come to predominate the Respond planning layer. For example it can deal with unexpected variations in customer orders (within probabilistic preplanned thresholds). Or identify odd demand behavior to make better decisions to reduce risk. Algorithms could analyze data patterns from smart connected devices, such as those on shop floors, in warehouses and with customers and distributors, and send a correlation, corrected plan or new plan back to help with respond capabilities or automatic replenishment.
This approach requires a unified data model to support such an ambitious supply chain ecosystem. It can be thought of as the “supply chain digital twin“. This is a digital representation of the current state of the supply chain (events, inventories, open orders, shipments, plans, scenarios and so on) and, as such, allows a company to examine (through the use of appropriate analytics) how the supply chain will respond to certain events and decisions. The best digital twins leverage a cloud environment like Azure and it’s PaaS (Platform as a Service) that delivers scalability, interoperability, consistency and security.
Click below for a copy of a more in-depth whitepaper on the topics discussed in this blog as part of “digital transformation”.