Artikel

GenAI in Supply Chain – Hope or Hype?

Artificial intelligence (AI) has become a cornerstone of innovation across various industries, transforming how businesses operate. Within the AI spectrum, Generative AI (GenAI) stands out for its potential to perform a wide range of interactive tasks with minimal human intervention based on its Large Language and Action Models (LLMs & LAMs). Unlike narrow AI, which is designed for specific functions like language translation or image recognition, GenAI can adapt to new tasks without extensive reprogramming. This capability makes it a promising technology for complex systems like supply chain management, where flexibility and efficiency are paramount. But does GenAI truly offer a hopeful future for supply chains, or is it just another overhyped innovation?

Maturity of current technologies in supply chain

The supply chain sector has embraced various advanced technologies at different maturity levels. Machine learning is becoming essential in demand forecasting, predictive maintenance, and logistics optimization. Companies like Amazon and Walmart utilize machine learning algorithms to predict customer demand, optimize inventory levels, and streamline delivery networks, leading to reduced costs and improved customer satisfaction. 

Blockchain technology has seen significant advancements over the past decade. Its unmatched transparency and security offer a tamper-proof ledger, valuable for industries like pharmaceuticals and food, where traceability is crucial for safety and compliance. Despite its potential, blockchain has not yet achieved widespread adoption due to scalability, cost, and interoperability challenges in global, multi-party supply chains. 

Robotics and automation have also progressed, with companies like DHL and Amazon using robots for warehouse management and order fulfillment. These systems enhance efficiency and accuracy in repetitive tasks. However, implementing heavily automated warehouses often requires significant investment and a complete redesign, resulting in most companies to automate only select operations, relying heavily on manual labor. 

GenAI use cases in supply chain

GenAI can enhance and expand machine learning functionality within global supply chain technology by broadening the range of data types used (text, audio, visual, structured, and unstructured data) and generating responses to problems or queries. 

Planning: Increasing accuracy with a broader information base 
GenAI’s ability to process multiple data formats can improve demand planning by incorporating additional information like market trends and geopolitical scenarios, enhancing historical data-based forecasts. 

Example: PepsiCo uses GenAI to review past sales data, analyze market trends, and consider external influences to predict future demand, optimizing production schedules. This enables PepsiCo to anticipate demand increases for seasonal items, maintaining sufficient inventory levels, minimizing waste, and avoiding stock shortages. 

Sourcing: Automating supplier negotiations 
For low-value, low-risk materials, GenAI can automate supplier negotiations, acting as a procurement agent to propose contracts. This reduces costs by simplifying and structuring tender processes. 

Example: Walmart employs a GenAI-based tool for autonomous supplier negotiations, securing better deals and streamlining the process. Many suppliers prefer negotiating with AI over humans. 

Producing: Increasing efficiencies and speed 
GenAI can create efficient production and maintenance plans, especially in resource shortages. It can also support in product design by generating concept, blueprint, or recipe variations. 

Example: Bristol Myers Squibb uses GenAI to design molecular glues, optimizing parameters like potency and selectivity, through a collaboration with VantAI. 

Distributing: Optimizing loading and routing 
GenAI can design new supply chain networks and optimize load and routing by analyzing load parameters, transportation constraints, weather forecasts, tariffs, regulations, and shipping fees. 

Example: While it's unclear if GenAI is used, UPS’s AI-based ORION system has saved about 100 million miles and 10 million gallons of fuel annually through route optimization. 

Is it the right time for GenAI in supply chain?

GenAI has the potential to address long-standing supply chain challenges, such as risk prediction and customer service automation. Its ability to analyze diverse data inputs and enhance human interaction makes it suitable for dynamic supply chains. However, its implementation should be approached realistically: 

  1. GenAI technology is still developing: Significant investment in infrastructure, data management, and skilled personnel is required to harness its full potential. Seamless integration into existing systems is crucial for success. 
  2. Widespread adoption is years away: According to Gartner’s Hype Cycle for Supply Chain Strategy 2024, GenAI is at the "peak of inflated expectations," with widespread productive adoption expected in five to ten years. Companies face challenges like data quality management, complex system architecture, and a lack of data science skills, particularly mid-sized and smaller companies. 

Companies should focus on addressing data management, architecture, and personnel skills before adopting GenAI. According to a 2024 Forbes article, around 90% of GenAI proof of concept pilots will not move into production soon. 

GenAI, as other data-driven technologies, essentially need to support the decision-making process in a company. As such, we suggest companies take a more holistic view and assess their data-driven decision-making capabilities and create a sustainable roadmap with interim objectives. Mastering other digital technologies such as data mining, robotic process automation, predictive analytics, or maintenance may be more beneficial initially. Specific GenAI use cases should be evaluated against conventional alternatives with a sound business case for further investment. 

Conclusion – Hope or hype?

While GenAI shows promise for enhancing supply chain efficiency and innovation, its current capabilities and widespread adoption are limited. Companies should first address foundational issues like data quality, IT infrastructure, and data science skills, incrementally improving digital capabilities. 

To learn more about preparing your organization for GenAI or exploring a tailored assessment of your supply chain’s readiness, contact us to navigate this complex landscape and build a sustainable digital transformation roadmap. 

Spörri, C.