Business disruptions brought on by the coronavirus pandemic — from empty shelves in grocery stores to long delays in e-commerce deliveries — have shined light on supply-chain weaknesses, and a growing opportunity for artificial intelligence.

Part of the problem has been the switch to just-in-time manufacturing (JIT), which has created lean supply chains that hold lower levels of inventory in order to remove the risk associated with overproduction and surplus. This strategy, borrowed from the automotive industry, has enabled suppliers to lower their costs through low inventory levels and reduced production costs. Yet, when something like a pandemic or a natural disaster occurs that creates a surge in demand, it’s difficult to ramp up production, or tap surplus supplies to fill the pipeline — especially when entire plants are shut down because of infection. What’s required today instead is a new model that enables real-time demand manufacturing over JIT.

AI-Driven Tool Kit 

More suppliers are turning to various forms of AI to better address the challenges of supply-chain disruption now and into the future. Consider the following five applications:

• Predictive analytics. So that companies don’t get blind-sided by changing market, economic or consumer changes, AI-based predictive analytics are enabling accurate forecasting by analyzing patterns in historical data. It uses data mining, statistical modeling and machine learning to enable huge data sets to predict future outcomes. For example, a retailer can use it to determine the likelihood that specific items will be out of stock and when, or the likelihood that a consumer will still buy the Brand X of paper towel if production halts on Bounty. It also could analyze suppliers to determine which ones will prove most reliable in an emergency situation.
• Deep learning. Having visibility into a store location remotely has become crucial during the pandemic. Today, video surveillance, combined with deep-learning-based solutions, can help managers determine if safety protocols are being followed, such as cleaning down POS systems, insisting on mask use and social distancing. Also, these types of AI-based video surveillance can help with store audits for inventory management to determine where the empty shelves are located and what’s not selling. 
• Warehouse robotics. Robots on warehouse floors, picking and packing items are substantially increasing the speed and efficiency of warehouses. During the pandemic when there has been a shortage of human labor because of sick workers or those with compromised health conditions that forced them to remain at home, robots have picked up the slack. 
• Robotic process automation (RPA). Repetitive tasks, such as invoicing, order processing, data entry and other administrative tasks can be a major culprit to supply-chain disruption. RPA has enabled many companies to automate those tasks, and free up human employees to handle more strategic tasks, as well as deal directly with partners, customers and others. For example, RPA can be used to process purchase orders, identify required inventory levels and match them against actual stock — all without relying on human intervention, except to address exceptions. 
• ComputerVision. This AI-driven form of image detection can be used in transportation and logistics, to help identify high-traffic areas and help plan the best trucking routes. ComputerVision-based algorithms analyze digital images or video from satellite imagery to spot and count, for example, cars and buses in certain areas and help truckers avoid those areas. In other applications, it can be used to determine where there may be damaged train tracks that could impede the smooth supply of goods by train. 

MicroWarehouse Approach 

In addition to automation, what will become a hallmark of future supply chains is warehousing and storage located closer to customers. Even from the likes of Amazon or Walmart, big warehouses will be replaced with many smaller ones — serving a customer base within two miles of it. Instead of having a few large warehouses, companies can have many to decentralize the distribution process, get goods to retailers and consumers faster and even boost sustainability practices, with less air and highway emissions. 

In fact, the supply chain of the future will very well be comprised of fewer human workers; greater AI-driven automation for managing and predicting inventory needs, processing data and handling back-office tasks and warehouse operations; and new delivery channels, including drones, for safer, contact-free home deliveries. 

The supply chain of the future will need to get over key hurdles to become a reality. Interoperable, integrated systems will need to share data across the supply chain to be effective. But companies will need to be open to this interoperability to succeed. Another challenge is the fear of job loss that automation will bring. Companies must dispel these fears by opening up opportunities for workers to take on new, more strategic roles, offering training and continuing education support; and reinforcing the benefits of AI, without losing sight of the unwavering edge humans will always have over AI. 

COVID-19 has shone a spotlight on the frailties of the supply chain, but through its lessons we are beginning to reimagine the supply chain of the future, which will be driven by AI and empowered by human ingenuity.

Carlos Melendez is chief operating officer of Wovenware.