AI Alone Cannot Future-Proof Supply Chains Against CrisesAI Alone Cannot Future-Proof Supply Chains Against Crises

Over Christmas, the Guinness shortage – or “stout drought” – made headlines as pubs were forced to ration pints amid supply shortages. Gen Z were blamed due to the popularity of the “splitting the G” trend on TikTok, however, in cases like this, we should be looking at supply chain planning systems and whether they are properly integrated with the operation.

It’s hard to quantify missed opportunities, but let’s look at the facts. Guinness owner Diageo recently announced that some 34 million pints were sold in pubs over the Christmas break, a reported increase of 2.5% year on year. While that’s certainly impressive, what would have the impact on sales been had the brewer been able to keep pace with the 19% increase in demand that was being reported as early as November?

This shortage is part of a larger trend of big brands failing to predict and respond to shifts in consumer behavior and supply chain disruptions. From KFC’s 2018 chicken shortage to PlayStation’s struggles to meet demand during lockdowns, it’s clear this is not a one-off issue.

Traditional forecasting methods have long struggled when it comes to gauging consumer demand and it remains one of the biggest challenges in industry. Undercook it and you end up in a situation like Diageo; Overestimate demand and you can be sitting on excess stock that you can’t shift.

Related:Why AIoT Is the Next Big Thing: AI and IoT in Harmony

In today’s market, which can often feel more volatile than ever, modern solutions combining  AI, simulation and automation can make a significant impact. Through these solutions, businesses can improve adaptability by modeling scenarios and rehearsing responses better than ever before.

Going Beyond Traditional Forecasting With Predictive AI

Businesses are limited in their ability to respond to disruptions quickly, with the average response to supply chain disruption taking two weeks. Over the course of a decade, this can cost a business 45% of one year's profits.

Traditional forecasting and also many AI approaches rely on historical data to understand patterns. While extremely useful for planning around known events, its limitations are exposed when it comes to unprecedented – yet plausible – events. In the case of Guinness, it appears planning tools struggled to forecast the extent to which a social-media trend, amplified by a marketing campaign and layered across many other complex demand signals, would impact sales.

When combined with simulation technology, millions of variables can be processed simultaneously, from social media trends, weather patterns and economic indicators, to reduce forecast error by 20% to 50%. Approaches that combine AI with simulation to enable us to look forward by creating synthetic datasets describing future events, so we can prescribe scenarios and help businesses prepare in the best way possible.

Related:Fourth Industrial Revolution: How AI Agents Are Transforming the Future of Work

Relieving Peak Pressures With Robotics and Automation

It is well established that one of the main causes of the UK’s productivity challenge is due to the relatively slow rate at which robotics and AI are being deployed. On average, UK workers contribute less economic value per hour compared to countries like France and Germany, partly due to limited investment in technologies that can automate repetitive tasks and create higher-value work opportunities.

The recent stout drought is a good example of how this issue is playing out in supply chains, where the adoption of automation and robotics has the potential to provide businesses with new levels of agility that can be activated quickly and efficiently.

Alleviating peak pressure is one example. Last year, ASDA invested in 164 robots to accommodate expansion and rise in demand. As a result, the product picking rate of the existing system was doubled, and the accuracy rate of deliveries increased to 99.8%.

While automation is key to enhancing supply chain agility and efficiency, it won’t fully deliver on the promise of adaptive supply chains alone. Robots must operate within a wider network of processes, systems and flows to be fully effective. AI’s role is particularly important here, helping to monitor processes and optimize the deployment of automation while identifying and addressing unexpected kinks that might result in unintended consequences elsewhere in the supply chain. This is key when rolling out at scale, where AI-powered simulations can help guide ROI decisions and understand the impact of new systems across multiple sites.  

Staying One Step Ahead With Simulation

Central to success is the use of digital twins combined with simulation – AI-driven replicas of business environments that can identify more obvious bottlenecks but also subtle interdependencies that human insight might miss.

It was only through simulation that one manufacturer I spoke with recently discovered that their warehouse automation systems – which had seemed optimal when viewed in isolation – actually created ripple effects that reduced efficiency by 23% during peak periods. This kind of insight is impossible to gain through traditional analysis methods.

While AI adoption is often seen as a heavy investment, modern solutions are becoming more accessible. Cloud-based AI platforms and modular systems allow businesses to start small, prove ROI, and scale up gradually.

One mid-sized beverage distributor implemented a digital twin for under £60,000 and achieved ROI within six months through better inventory management. This approach not only enhanced forecasting accuracy and resilience but also helped save money by reducing disruption risks and finding the quickest path to value.

The Guinness shortage represents more than a temporary inconvenience; it’s a wake-up call for supply chain modernization.

As market dynamics become increasingly complex, the gap between AI-enabled and traditional supply chains will widen. By integrating AI-powered solutions such as simulations, digital twins and software to optimize robotics, businesses can reconfigure their operations and be confident that they are ready to accommodate a wider range of potential scenarios.

Those who delay modernizing their operations risk being left behind in an increasingly competitive global marketplace.

This article first appeared in IoT World Today's sister publication AI Business.