Supply chains have always been fragile, and nothing exposed it more than the COVID-19 pandemic. Empty shelves in grocery stores and delays in deliveries have highlighted supply chain weaknesses.

The pandemic wrecked the popular just-in-time manufacturing (JIT) model. JIT’s lean supply chains and low inventory levels eliminated the risks and costs associated with surplus. But businesses ended up with a stock-out situation as soon as the pandemic disrupted the supply chain.

Accenture’s July 2020 CxO Pulse survey reveals three out of every four supply chain executives are rethinking their supply chain processes and operating models to improve resiliency. Businesses turn to Artificial Intelligence (AI) and analytics for solutions. Here are six ways Ai and analytics can make supply chains more resilient and reliable.

1. Predictive analytics

Predictability makes the life of a supply chain manager easy. Predictive analytics give managers insights into the ever-changing economic, environmental, and market conditions. These insights enable them to make prompt changes and avoid disruptions.

Artificial Intelligence (AI) powered predictive analytics tools enable accurate forecasting using historical and current data. These tools apply data mining, statistical modelling, and machine learning to identify patterns.

AI-enabled predictive analytics:

• Improve demand forecasting. The insights help managers determine region-wise optimal inventory levels to right-size inventory levels. For instance, a store could stock more umbrellas and raincoats, if the analytics predict rain.
• Avoid stock-out situations. Retailers deploying predictive analytics on sales data predict when specific items go out of stock. They may also identify supplier lead times to determine supply reorder levels. They may also short-list reliable suppliers who deliver fast.
• Boost sustainability. Optimal stock levels reduce waste, save inventory storage costs, and minimise cargo movement. Suppliers do not have to replenish stock frequently, and sellers do not incur expenses to store dead stock. Carbon emissions reduce across the supply chain.

2. Unified demand planning

Understanding the mind of the individual end-customer to meet their demands is challenging. In today’s environment, customer sentiments are fickle, and expectations change rapidly.

Convention analytics rely only on internal sales data. Such data is helpful but incomplete for a complete analysis. AI-powered solutions capture relevant external data and make explicit the influences that shape demand. Advanced AI tools enable predicting demand before customers decide to buy the product.

Unified demand uses Big Data and AI to integrate internal and external data cutting across processes and functions. The data processing takes place in real-time to predict live demand. Such a unified, live view of demand enhances supply chain decision accuracy.

Advanced algorithms leverage:

• Internal data of the company, such as supply chain and trade data
• Customer data, including consumption data
• General data such as macroeconomic factors, market sentiments, and weather patterns

A deep and real-time analysis of insights enables accurate forecasting of customer demand. The business could fix optimal shipment quantities at a location level. Companies become prepared to meet changes in demand without facing disruptions.

3. Advanced scenario modelling

The disruptions caused by the COVID-19 pandemic have brought to centre stage scenario modelling.

The primary tool for scenario modelling is digital twins. A digital twin is a virtual supply chain replica. It co-opts assets, warehouses, inventory positions, and logistics and material flows, virtually. AI tools simulate these virtual digital twins to find out how different complexity drives value loss and risks. Simulation of these digital twins unearths potential vulnerabilities. Enterprises can identify their vulnerable areas and optimise them.

For instance, Accenture and MIT have used digital twins to create a supply chain stress test. The test uses AI-powered modelling to assess potential operational and financial risks. It also assesses the impact of disasters and other disruptors.

Supply chain analysts may create digital twins for the end-to-end supply chain or specific functional areas. Either way, the overall supply chain efficiency and resiliency improve.

Enterprises may use scenario modelling to

• Address the weak links in their supply chain and improve resiliency. Optimise cost.
• Improve customer service.
• Reduce carbon footprints to meet sustainability targets.

4. Deep learning

Combining video surveillance with deep-learning-based solutions offers enhanced visual insights. Deep visibility enables:

• Better adherence to safety protocols.
• More efficient inventory management. Visual-based store audits identify empty shelves and the accumulated stock of slow-moving items. Bots enabled with computer vision automate repetitive tasks, such as real-time inventory scanning.
• Improving quality assurance. AI-powered computer vision systems enhance the quality assurance of finished products. Automobile majors such as BMW use computer vision to scan car models as they move on the assembly line.

ComputerVision, or AI-driven image detection, identifies high-traffic areas or route disruptions. Supply chain controllers may use deep insights to plan the best trucking routes. For instance, the algorithms analyse satellite imagery to count the number of vehicles on a specific path. If the number of vehicles is high, the system alerts truckers to avoid those routes or areas.

5. Smart warehouses

Warehouses are central nodes of any supply chain system, but also major supply chain bottlenecks. Applying AI and analytics to warehouse operations and designs removes these bottlenecks.

• Deploying robots to pick and pack items improves the speed, efficiency, and accuracy of warehouse processes. Unlike humans, robots do not suffer from fatigue or take leaves.
• AI-enabled technologies, such as cobots, drive efficiency, productivity, and safety in warehouse management.

A MicroWarehouse approach decentralises warehouse operations. Giant, centralised warehouses are disruption prone. Resilient supply chains rather feature hundreds of smaller warehouses and storage capacities. These decentralised warehouses become critical nodes to deliver goods faster and safer. It also boosts sustainability with less air and highway emissions. Powering such decentralised micro-warehouse operations is the free flow of data. Robust Big data analytics tools to divert the right parcel to the right warehouse.

6. Supply chain automation

Several large companies have taken the lead in supply chain automation. Companies such as Amazon, Nuro, and Tusimple have invested extensively in autonomous trucks.

A significant and often underestimated cause of supply chain disruptions is paperwork. Supply chain companies focus on cargo movement, and give administrative work the short-shrift. But any delay or laxity in administrative compliance or paperwork can disrupt cargo movements. Tasks such as order processing, invoicing, payment follow-ups, and data entry have a big share in supply chain disruption.

Automating administrative tasks using robotic process automation (RPA) puts these tasks on auto-pilot. AI-powered decision making processes orders, identifies required inventory levels and matches them against the stock. Everything takes place at the right time. The human workforce only has to handle exceptions.

Automating the viable areas of the supply chain:

• Improves efficiency and cuts costs, boosting both profits and sustainability.
• Reduce incidence of delays
• Frees human employees to handle more strategic tasks, such as dealing with customers and value chain partners.

7. Enhanced supplier relationship management

Managing demand is only half the battle. The other half of the supply chain battle is similar visibility in the supplier base.

Weak supplier relationship management is the root cause of many supply chain delays and disruptions. Lack of collaboration and integration among suppliers undo supply chains during times of crisis.

AI makes supplier relationship management more robust, consistent, and efficient. AI-enabled software shortlist suppliers based on purchase history, pricing, infrastructure, and sustainability. Analytics tools collect supplier data to assess performance and rank suppliers accordingly. Enterprises may capture data across supply chains and apply robust AI models to gain real-time supplier insights. Enterprises could:

• Analyse the suppliers’ financial position. A supplier with a weak financial record could risk the supplier’s ability to fulfil the order. Even a small disruption could lead to stock-out of vital raw materials.
• Ensure the suppliers’ carbon footprints meet agreed-upon targets.

Analytics, AI and the cloud enable companies to monitor and respond to disruptions within the supply chain. During the COVID-19 pandemic, most enterprises were effectively flying blind. Lack of insights into the supplier network meant they could not identify the shut-down suppliers and suppliers still processing orders. Even a minor delay from one supplier can have cascading effects that cause huge downstream complications.

Better information leads to better decisions. The supply chain of the future uses technology to gain more knowledge and eliminate hurdles. They will use interoperable, integrated systems that share data across the supply chain. AI-powered big data makes the supply chain agile, resilient, and sustainable. So far, businesses have only scratched the surface of the possibilities offered by analytics and AI.