Thanks to the growth of generative AI, a truly autonomous supply chain may be closer than we think. Here's why.
Ask any supply chain professional over the last year and they'll tell you that their company wants to reap the results of generative AI. Research by EY backs this up, finding that nearly three-quarters (73%) of supply chain and operations executives are planning to deploy GenAI. However, just 7% of them say they have successfully implemented the technology.
Making the technical leap from proof-of-concept to GenAI at scale is challenging. This becomes even clearer when you consider that supply chain operations everywhere struggle with data quality, organizational readiness and volatility -- both internally and externally. However, organizations who have invested in AI early on have, at least partially, broken through these barriers. A 2023 survey from McKinsey found that supply chain and inventory management were two areas that report meaningful revenue increases through AI.
To take advantage of this however, it must become simpler for teams to integrate GenAI in their everyday workflows. Here's how.
Accurate, proactive planning
Supply chain success is built on a foundation of smart decision-marking. However, without a base of historical business knowledge, supply chain planning leaders are left to manage important lead times and inventory based off gut feelings rather than accurate supply and demand data. The resulting guess work impacts lead times and ultimately affects customer satisfaction.
"Data integrity is one of the most powerful components to consider as we move towards the era of autonomous supply chains -- it's essential to enable a seamless end-to-end process across the entire supply chain," says Mindy Davis, global vice president, product marketing for digital supply chain at SAP.
Many companies on a digitalization journey may have eliminated most of their paper-based systems to gain better control over their supply chain, but they haven't fully incorporated these types of analytics into their decision-making processes. EY's research found that even for organizations using GenAI in their supply chain, only 50% have achieved end-to-end visibility. "It may sound antiquated, but quite frankly, digitizing paper-based systems is the first step to establish a digital foundation so you can access comprehensive data that impacts your supply chain," says Davis.
AI also proves to be a powerful tool for planners to get a leg up and bridge this gap. With verified and consolidated data, supply chain teams can train AI models to help accurately predict future lead times or track the status of shipments in real-time. Accurate lead times mean teams can deliver the right products at the right time and keep customers satisfied. As companies move toward an era of autonomous supply chains, an AI solution integrated into their ERP can help supercharge business decisions.
At SAP, Davis and her team leverage business and financial data found in the company's ERP solution, SAP S/4HANA, to help planners using their other applications, like SAP Integrated Business Planning, to make more informed decisions and accurately predict lead times. Then, using SAP's AI copilot Joule, they can get better insight into the variables or constraints facing their inventory and solicit recommendations to be more proactive in their planning.
"We're envisioning a path characterized by technological, procedural and data enhancements that will propel the supply chain into an autonomous era, where supply chains operate with minimal human intervention," says Davis.
Efficient, error-free manufacturing
In today's supply chain environment, there really is no room for disruption -- be it labor shortages, geopolitical strife or malfunctions within manufacturing. To keep up with demand, supply chain teams are focused on continuous improvement and finding ways to remove the burden on expensive manual labor in favor of automated, digital solutions.
When faulty products come off the production line, it must be addressed quickly. AI can accelerate the resolution process faster than human labor in many instances -- preventing production standstills and even catching errors before they occur. Engineers who are creating a product can lean on these insights too, using AI to assess all the errors that have happened in the past to make sure that they don't happen in the future.
But AI doesn't just improve error resolution, it can transform the first phases of production as well, such as eliminating redundant tasks like tagging data on product visualizations. It can even make designs more efficient, developing, enhancing and customizing recipes for products while supporting product compliance and sustainability.
Powerful, predictive maintenance
Manufacturing the millions of products that move throughout the supply chain starts with the machinery used to produce them. This "up" time -- hours when equipment is in action -- is the backbone of effective operations. When those components fail or reach the end of their lifespan, that has an enormous operational and financial impact, shifting budget, influencing payment terms and mitigating cash flow.
Regular maintenance on essential equipment is key to keeping the supply chain moving. But monitoring wear and tear to catch issues before they even happen is even better, and increasingly possible thanks to AI.
Through camera footage and visual inspections, AI models can help detect errors, faults or defects in equipment before they happen. If the technology identifies an issue -- or predicts the need for maintenance -- teams can arrange for a technician to perform repairs. This predictive maintenance minimizes unplanned outages, reduces disruptions across the supply chain and optimizes asset performance.
Swiss Federal Railways (SBB) is piloting this capability through SAP to help get passengers where they want to go on time and safely by examining a small but critical piece of railway infrastructure: the pantograph. This component mounted on the roof of an electric train collects power through contact with an overhead line.
"What many companies will do is implement generative AI tools for small uses cases,"
says Davis. "Implementing use cases that show initial success helps show the potential of this technology for the supply chain, and then you can consider scaling across your organization."
Using AI, SBB can assess the pantograph's thickness and conductivity to determine when it should be replaced or repaired. The railway is even adapting this process to examine other equipment, including doors, chairs and food equipment onboard.
These AI-enabled inspections are part of recent updates to SAP Asset Performance Management. For supply chain managers, the condition data is compiled through Joule and connected to S/4HANA, contextualizing the information and connecting the business functions. This way, they can monitor these scenarios and get AI-powered recommendations on the best next steps.
All of these innovations are steps along the path toward a truly autonomous supply chain. While first-movers have integrated AI into their process and found positive results, it remains a challenge to integrate these solutions across an entire organization. As companies continue their digitalization journey, finding applicable use-cases for AI will be important to accelerating their progress and seeing real-world benefits of the technology.
