The pandemic was a confusing and odd moment for demand-driven digital supply chain proponents. Two decades of automation and analytics-led supply chain digitization had left supply chains so lean that system-level shocks resulted in broken supply chains. These contributed to rampant inflation, layoffs, lower manufacturing output, social unrest, and empty store shelves. Digitization in every sector got a boost during Covid but supply chains had an existential crisis. Lately, a new supply chain reality is rising from the ashes of demand-driven supply chains and it revolves around AI.

AI in supply chains will affect all their aspects including inventory management, logistics, processes, and business decision-making. Technological advancements in supply chains will make them seamless, reduce costs and drive efficiency in business processes.

The challenge is to frame AI’s role in supply chains well, invest in the infrastructure personalized for companies and then ensure superior first-party, second-party and third-party data streams to power analytics. Understanding how automation, analytics and risk change in the AI age will be key to transformative supply chains.

This will not be neither cheap or easy.

Systems: AI systems are expensive to install and run. AI development costs for large supply chains, aiming for high algorithm accuracy with high data throughout require infrastructure optimization, data engineering, modelling and integration, security, and artificial intelligence management and control efforts. Many AI systems are cloud-based but can require customized hardware to run and there can be significant costs of bandwidth. These can all add up.

People: Hiring people to manage the AI in the supply chain as well as training existing resources is expensive.

Data, data, data: The state of AI projects is dismal. Successful projects that deliver intended benefits are less than 20% by some measures. A large reason for this is that everyone wants to do data modelling, but no one wants to do the data engineering. Without high quality and clean data, algorithms fail. This requires not only investments in people and processes, but also a company-wide data strategy. Data is hidden in silos with legacy knowledge as a barrier to entry and jealously guarded by line of business stakeholders. It is a singular and often insurmountable challenge to get it out.

Is it worth it?

Some early adopters have seen a decrease in logistics costs of 15%, an increase in inventory levels of 35%, and a boost in service levels of 65%. Most saw a revenue increase of over a third and also saw a considerable revenue bounce. These are transformative numbers. The question is not if we should embrace AI in the supply chain but rather how we do it in such a way to ensure the success of the initiative. This requires preparing not just the supply chain function but the entire organization for change.



What will AI change?

While targeted supply chain projects are smart as a proof of concept to demonstrate viability to the board and secure additional funding, the full value from AI will be realized when AI is embraced across the supply chain and strategy should be made accordingly. AI in Supply Chain affects five key supply chain areas.

Inventory management: Inventory management is a mercurial discipline. Masters of inventory management need to see both the forest and the trees, simultaneously looking at third-party data on industry and sector trends, idiosyncrasies of customer behavior, historical data, and the company’s growth initiatives. Effective inventory management can save the annual balance sheet. One major slip-up can take it into the red. AI was built to manage exactly such complexity. AI platforms and tools can take historical data and add to that real-time warehouse data, industry inventories, and even news on key competitors and ensure supplies never reach critical levels while cash flow remains optimized. Inventory replenishment, forecasting time of arrival and safety stock management are some metrics that see improvement with AI.

Logistics: Transporting, warehousing and storage runs companies, countries, and the world. AI sensors can address use cases like reducing cold chain issues and tracking multiple shipments to ensure accuracy and give visibility. This requires combining IoT, Edge computing, cloud, and AI. AI route planning makes the supply chain greener and saves costs. Smart warehouse systems make operations cheaper, faster, and more efficient. In one case a massive distribution warehouse for an e-commerce platform company with over a hundred self-charging, Wi-Fi-powered robot vehicles tripled worker productivity.

Process automation: AI can make better more efficient workflows. It can eliminate redundant, repetitive, manual processes. Supply chains, especially global supply chains are heavily dependent on documentation. On one hand, AI can be used to digitize existing and historical documents to gather high-quality data and on the other, the new digital workflows that replace them are far less erroneous and provide greater productivity.

Decision making: With the right company-wide data strategy in place, business managers will get more smartly curated high-quality data in a timely and relevant manner. They will have superior planning (long-term decisions) and smarter reactions to markets and events (short-term decisions), so better decisions are faster every time.

Risk: Risk in the supply chain clearly needs revisiting as the pandemic revealed this to be a fractured promise. However mitigating risk using AI is now the baseline. The key here is data quality. The organization needs to control data aggregation and utilization inside its organization, so it consistently uses clean high-quality data. Critically you need to match internal data with superior external data. The biggest supply chain risks often come from suppliers and their suppliers. High-quality external data with tools that can be used to flag supplier risk and ensure governance and compliance with company and country regulations are an essential part of the modern supply chain.

Leaner supply chains were more efficient but vulnerable. With AI embedded end-to-end in the supply chain, we can build smarter resilient supply chains. As Supply Chains move from descriptive intelligence by continuously learning on high-quality data, they will become predictive and prescriptive running 24/7/365 with certainty.

D&B takes a lead

As a global leader in introducing targeted and tech-enabled solutions, Dun & Bradstreet has taken a lead in integrating artificial intelligence in a diverse range of its upcoming products. Our products, driven by AI-powered data and insights, will assist businesses in decision-making and risk mitigation.

To learn more about our AI-enabled solutions, stay tuned to