Examining Artificial Intelligence’s Role in Global Supply Chains

Download this article in PDF format.

A branch of computer science focused on building smart machines capable of performing tasks that typically require human intelligence, artificial intelligence (AI) has made its way into the mainstream over the last few years. From Google’s AI-powered predictions to the dynamic pricing algorithms that Uber uses to Siri’s voice-activated personal assistant capabilities, AI has become part of our daily lives (whether we know it or not).

In supply chain, AI is being used to forecast demand for products, manage autonomous transportation, improve routing efficiency and chat with customers online, among other things. And while Secondmind says that the adoption of AI to improve supply chain decision-making remains at a relatively low 12%, the use of AI in that context is growing rapidly.

“AI has long promised to have a transformative impact on business,” the company points out in its report. “The supply chain, with access to more data than ever before, and with key forecasting and planning decisions ripe for improvement, stands to benefit handsomely.”

Frustrations Mounted

In a survey based on input from 500 supply chain managers and planners in the U.K., U.S.A., Italy, Spain, Germany and France, Secondmind found that while 90% believe AI will help inform better decisions in the future, 82% experienced frustrations with current AI systems during the COVID-19 pandemic.

“Organizational challenges and the longstanding issue of data quality were highlighted as key issues which led to expensive mistakes, less accuracy, and slower decision-making during rapidly changing market conditions,” it says. But that’s not to say those who are testing the waters of AI are ready to give up on it; quite the contrary, in fact.

“The message from planners and managers is clear: people need AI, and AI needs people,” Secondmind points out, noting that the ability to feed in their insights, teach AI when data is unreliable and modify forecasts based on their own judgement were all cited by planners and managers as desirable capabilities in AI systems.

During its research, the company also found that:

• AI is gaining more followers. It might not yet be widely adopted in the supply chain, but planners and managers who use AI are strong advocates for its ability to enhance their decision-making. Nearly all believe that AI will transform supply chains for the better in the next five years. In fact, 99.8% of those using AI-powered software and tools said it helped them do their jobs more effectively, and 74% said they can make better decisions if forecasts and recommended decisions are produced using AI alone.
• Data and organizational readiness are lacking. According to Secondmind, a lack of reliable data, the problems with historic data in times of change and the need to revert to human analysis were all blockers at a time when accuracy and speed were of the essence (i.e., during the depths of the COVID-19 supply chain crisis). “The problem of sparse, incomplete or unreliable data has been a longstanding challenge in the supply chain and one which isn’t going away anytime soon, suggesting that AI systems need to be able to work effectively and efficiently in such environments.”
• Companies need to be able to think faster on their feet. As AI-powered systems struggled to cope with the disruption caused by the pandemic, Secondmind says significant time was spent manually analyzing and interpreting data, leaving supply chain planners unable to respond to changing market conditions quickly enough. “Instead of making invaluable contributions on key activities, such as scenario planning for future black swan events, risk assessments, and more in-depth analysis using their experience,” the company states, “supply chain managers and planners were instead firefighting problems caused by system failures.”
• AI is not a silver bullet. According to Secondmind, supply chain planners and managers want AI systems that enable greater collaboration in decision-making (i.e., AI that “learns from their experience”). Others want more context for recommendations (i.e., which circumstances impacted a forecast, or how confident the AI system is in its predictions) in order to make better, more informed decisions.

The Human Element Counts

As AI continues to evolve, expect to see it playing a bigger role in supply chain management, albeit in tandem with human capabilities and expertise. “These insights serve as a timely reminder of the importance of human intuition and experience in the decision-making process,” Secondmind adds. “An over-reliance on data-driven insights to inform decisions will fail when data is sparse or incomplete, unless the AI systems are built to work in such environments.”