AI for Supply Chains Needs Cause-and-Efect Reasoning
Forecasting is at the heart of managing an efficient supply chain. Inventory management, especially the common challenge of inventory taking for sustained periods, has made intelligent forecasting more important than ever to enable our highly complex supply chains to function in real time and meet customer needs. However, the current state of the art in machine learning relies on past patterns and correlations to make predictions of the future—making it prone to failure amid shifts in data distribution.
The starting point to correct this is to shake off the misconception that machine learning is synonymous with Artificial Intelligence; the real AI revolution only begins when machines can learn like scientists – look at causal factors, as well as data, and make reasoned judgments for true intelligence in decision making. Today’s learning machines have superhuman predictive ability, but are not particularly good at causal reasoning, even when fed ever-increasing amounts of historical data to satisfy their appetite for finding and exploiting statistical regularities.
It is largely true that supply chains are facing an inevitable transformation through AI. However, AI’s focus so far on data has missed the fundamental problem: the world is changing very quickly, and basing actions solely on past data can lead to sub-optimal decisions. However complex available methodologies may be, they are only capable of inferring correlations. For outcome-focused decision making, machine learning needs to merge with, for example, “domain” expertise from humans to make more sophisticated ML algorithms.
The need to understand the cause and effect of possible actions to influence desired outcomes has long been understood in fields such as economics and medicine, but it has only recently begun to emerge in industry, let alone in the supply chain. When the causal drivers of supply or demand in the world change, even sophisticated curve-fitting models can make worse decisions than flipping a coin. It’s not just a matter of a data scientist retraining a model to reflect external changes; the model still works with static – albeit revised – data.
Progressing from predictive to prescriptive solutions to supply chain problems requires more than data scientists; it needs a holistic approach to building systems. At an extremely granular level, this involves taking data about timeframes and products, shipping distances and times of manufacture, and using it in knowledge and context about prevailing external factors. Currently, decisions are largely optimized at a “macro” level. It is not uncommon for a large enterprise to use only three machine learning models to address 10 million possible permutations of timeframe and location for shipping. However, by building in causality, it is possible to reduce waste, help the environment and improve profitability.
The next generation of AI will deliver KPI optimization platforms that look at a business as a whole, including understanding causal elements, to make critical business decisions. Be warned: the technology won’t be able to be downloaded from the open source community to be bent into shape by a few data scientists; it will involve “thinking” at a much higher level of automation.
Causal AI enables visibility of the entire supply chain to quickly understand and act to mitigate delay. Next-level AI isn’t about being satisfied that predictions are right. It asks: are we making the right decisions? Can we deduce the impact of our decisions? Do we know the cause of our outcomes? After all, KPIs and ROI are outcomes of decisions, which require a causal element. Supply chains are influenced by numerous external factors: relationship management; regulatory environment; operational risk; expert judgement; budget constraints; business context. Improving On Time, In Full (OTIF) service levels to a significant degree will be achieved faster with a full consideration of causal factors, as well as the use of random AI to assess what-if scenario and optimization planning.
So how can this work in practice? Through causal discovery and inference algorithms, millions of data features are defined and connected—not just statistical, but also causal relationships. Supply chain knowledge is embedded, allowing subject matter experts to inject more precision into a causal graph. Since supply chains are generally very complex, this domain element is essential. The combination of top-down human expertise and bottom-up data discovery is very powerful.
Specific intelligence on materials and shipping, supply and demand, production and purchase orders, or sales activity and orders are all areas from which it may be possible to identify factors conspiring to cause delay and friction in the chain. When a combined approach is adopted to achieve next-level causal AI, it can be deployed in numerous use cases, for example a process of root cause ID to correct sales order process delays. This root cause can be used within a decision application capable of providing actionable recommendations for business users and domain experts. In case of such delays, it can consider the impact of capacity optimization to identify the top five centers in a network for processing an order. This can be done programmatically, with outcomes to be tracked, reducing time and cost.
AI has been touted as a transformative development for the supply chain for some time. McKinsey estimates that organizations worldwide could earn between $1.4 and $2 trillion in revenue by using AI in manufacturing and supply chains. Nevertheless, in reality, a level of machine learning capable of truly optimal decision making is only now emerging. Using cause and effect, a new category of intelligent machines that reason as humans do will become a revolutionary tool for solving real-world challenges. On a three-to-five-year view, the future for AI in the supply chain is very bright indeed.
Jerry Stephens is GM of Supply Chain Management at causaLens
