The most recent McKinsey Global Survey on AI shows a 25% increase in use by companies over the prior year, a trend that shows no signs of slowing. At this pace, have we entered the Age of AI? Do we even need supply chain planners, or can we turn out the lights and go home, because we now live in a world of autonomous planning? The short answer is that planners are more important than ever, and not just because the hype may be ahead of the curve. According to the MHI Industry Annual Report, AI and machine learning adoption in supply chains is only 12% today, but the curve is expected to rise rapidly to 60% within five years. More importantly, while AI may promise the world driverless cars, supply chains still need humans in the driver’s seat. Although human hands need to remain on the wheel, it is time for a shift change. We are driving on a different kind of supply chain landscape these days, which calls for a different kind of planner.
The Changing Supply Chain Landscape
In days gone past, supply chains were about moving “stuff,” so the employees who ran them had logistics backgrounds relevant for their highly operational roles. Over time supply chain has been given more power and responsibility but also faces more challenges. The need to meet ever-higher customer expectations at ever-lower costs has lengthened our global supply chains, from multiplying far-flung suppliers to directly into consumers’ homes. Supply chains have seen the need to digitalize their operations to increase efficiencies to meet growing demands. An additional pressure is greater scrutiny on the sustainability impact of supply chains. And now, in the wake of the global pandemic, supply chains also face keen expectations to be agile and resilient and able to deliver end-to-end visibility.
Demand for AI Leads to Talent Gap
Investing in digital transformation was already a priority for supply chains, and while investments may now be more targeted, they are not slowing down. AI and machine learning (the subset of techniques most in use in business today) figure prominently on the agenda, as companies want to improve forecast accuracy with demand sensing, predict yields and lead times, optimize inventory, get recommendations on how to respond to disruptions, and more. The promise of machine learning lies not only in improved intelligence in these key areas but in automating them for greater efficiency. I am bullish on making supply chains smarter and more efficient using machine learning, but I am equally convinced that a key to their success is the ongoing importance of the planner in the process.
The demand for AI and machine learning has led to a corresponding demand for talent, with LinkedIn reporting annual growth rates of 74% for these roles. According to the previously-cited MHI report, talent is already the top challenge for supply chain leaders. In a research report by DHL, they report that logistics jobs grew 26% in the last decade in the US. This is not a US-only phenomenon: in the United Kingdom, the number of postings for supply chain roles are 219% higher than other sectors but attract fewer applicants. Finding talent in supply chain is hard enough, but attracting AI talent to look beyond roles in startups and big tech and consider supply chain is even harder.
Automated Machine Learning Addressing the Gap
Fortunately, an emerging approach to AI called automated machine learning (or AutoML) is expanding access to these techniques to whole new categories of people without requiring scarce expertise. By automating many of the steps involved in building a machine learning pipeline, a business user without training as a data scientist can build many usable models, especially for areas like common forecasting tasks. This development is so promising that McKinsey predicts demand for what they call AutoML practitioners to be double that of data scientists. When AutoML is paired with tools to ensure interpretability of models to mitigate their black box nature, the combination can ease adoption and offer tremendous business value to a far wider range of companies, including their supply chains.
Will AutoML Turn Off the Lights?
While the hype may herald a lights-off approach aimed at obtaining a fully autonomous supply chain, it misses the ultimate value of AI and machine learning – to augment the human, not replace her, because there are still areas where she can do things math cannot. Machine learning can find patterns in data far beyond the cognitive capacity of humans, but since math learns from the data it is fed, it is at a loss when context changes. In the immediate aftermath of quarantine, data from April 2019 was no more helpful to machine learning models than data from March 2020 to predict demand for April 2020. Instead, in the face of this kind of disruption, a human able to interpret the context of what was happening, armed with a keen understanding of their company’s business, was far better suited to applying expert judgement for making critical decisions.
In addition, in many situations conscience needs to be applied, for questions like responsible sourcing or determining how best to allocate scarce supply, to name just two examples calling for ethics, still a uniquely human domain. Similarly, establishing and maintaining the kind of collaboration and relationships at the core of supply chain is also best left to us humans. We maximize benefits when we combine humans and machine learning, so each can offer their complementary strengths. Math still can’t tackle context, conscience or collaboration, so humans remain relevant to bring these critical capabilities to bear on today’s supply chains. Pairing these human capabilities with the pattern-finding and computational capacity of machines can lead to business benefits like reducing stockouts, decreasing inventory, improving on-time delivery, and building sounder production schedules.
Building the Right Kind of Talent Pipeline for Supply Chain
The logistics planner of yesteryear needed one set of skills, but we have different expectations for the planner needed to run today’s supply chain. She still needs to have business acumen and supply chain knowledge to understand her environment, along with strong communication skills. But while she doesn’t have to be a data scientist, she should have enough data literacy to be able to understand how and where machine learning can be applied from a business perspective. In addition to these technical skills, she needs to be able to derive meaning from context, apply conscience, and collaborate. In fact, instead of serving one silo in the company to plan for materials or demand, she could even be a network planner, with scope across the entire supply chain. As fast as the pace is in today’s supply chain, agility and an understanding of change management will serve her well. So rather than eliminate her role, AI and machine learning can augment the planner, so she can focus on bringing all these skills together, along with the power of the math, to drive today’s supply chain into the future.
Polly Mitchell-Guthrie is the VP of Industry Outreach and Thought Leadership at Kinaxis, the leader in empowering people to make confident supply chain decisions. Previously she served in roles as director of Analytical Consulting Services at the University of North Carolina Health Care System, senior manager of the Advanced Analytics Customer Liaison Group in SAS’ Research and Development Division, and Director of the SAS Global Academic Program.
Mitchell-Guthrie has an MBA from the Kenan-Flagler Business School of the University of North Carolina at Chapel Hill, where she also received her BA in political science as a Morehead Scholar. She has been active in many roles within INFORMS (the Institute for Operations Research and Management Sciences), including serving as the chair and vice chair of the Analytics Certification Board and secretary of the Analytics Society.
