Artificial intelligence in the supply chain

AI can help manufacturers and retailers increase forecasting accuracy and optimise replenishment. E-commerce pioneers such as Amazon and Alibaba are already using machine learning to predict future demand for products based on consumer behaviour and seasonal or environmental factors, allowing for automated placement of product into fulfilment centres.¹

While Unilever has developed an AI-powered customer connectivity model that enhances collaborative planning, forecasting and replenishment process that has achieved a 98% product availability in a trial with Walmart and reduced human effort by an estimated 30%.²

Similarly in Ireland, An Post, the national postal service has been exploring AI to optimise delivery routes and demand forecasting.³   In the agri sector, chemical companies have used similar models and environmental data to predict which regions will need particular chemicals and nutrients at which times, allowing for enhanced stock positioning and capacity planning.⁴

Computer vision-enabled AI can allow manufacturers to detect product defects in real time, enhancing fault detection and reducing costs from scrap and reduced yield. For example, Foxconn recently integrated Google’s Vision Inspection AI to its manufacturing process and has reported enhanced quality as a result.⁵

Tesla similarly makes liberal use of computer vision and AI in its ‘unboxed’ manufacturing method, enabling it to identify and correct defects in real time,⁶  whilst AstraZeneca claims to have ‘reduced manufacturing lead times from weeks to hours’ through smarter manufacturing processes.⁷

Similarly General Motors has used AI in the design for manufacture process to reduce component weights by 40%, while increasing strength by 20%.⁸

In Ireland, companies across a range of sectors, including manufacturing and resource management, are working with research organisations to test new methods of production, develop innovative materials and ultimately rethink how they manufacture for more efficient operations and greater product circularity.

Throughout the product life cycle, AI can accelerate adoption of circular practises. From virtual testing to reduce waste from prototypes and process optimisation through to enhanced end of life.

Revolve, a member of Circuleire - Ireland’s first national circular innovation network has created a technology-driven supply chain solution to match reusable auto part and waste materials with appropriate users, this helped An Garda Síochána, the Irish police force, to save the equivalent of approximately 38,000 kg of CO2 in 2022 by acquiring over 550 reclaimed vehicle parts of various makes and models.⁹

Dr. Geraldine Brennan, Head of Circular Economy and Circuleire Lead at Irish Manufacturing Research, noted, "Many companies are yet to understand the role of AI in enabling greater circularity and manufacturing efficiencies. AI can help companies to identify opportunities for better material and resource utilisation in manufacturing, while also supporting reuse and remanufacturing." 

AI provides manufacturers with a range of options to optimise warehouse operations, including inventory management, space utilisation, automated packing, and order tracking.

FedEx and DHL have already rolled out AI-enabled sorting robots, while DB Schenker is piloting computer vision and AI in its warehouses to offer real-time pallet tracking and ultra-accurate dock-to-stock cycle times for inbound pallet management.^{10, 11, 12}

Mengniu Dairy, in China, has used AI to automate supplier order scheduling and vehicle dispatching, increasing inventory turnover by 73% and operational efficiency by 8%.¹³

Patrick Corbett, Managing Director, DHL Supply Chain Ireland, noted, “As the demand for faster and more accurate fulfilment has grown, business needs to leverage more advanced technology to scale operations and improve overall efficiency. Analytics and AI provide more precise forecasting, stock holding, pricing, and identification of slow-moving items, which helps to reduce overstocking and stock-outs, and ensure the right products are available at the right time. With the use of AI throughout our customers supply chain, we’re aiding in developing end-to-end automation capabilities that give our customers full control and flexibility over their logistics and supply chain operations.” 

AI, coupled with appropriate equipment sensor data, can autonomously predict when machines are likely to fail or require maintenance, which reduces unplanned downtime, extends equipment lifespan, and reduces maintenance costs by flagging issues before they lead to significant breakdowns.

GE, Bosch, and Rolls Royce are just some of the major engineering companies already using such technologies to analyse data from thousands of in-service units to detect anomalies, facilitate proactive maintenance and reduce lifecycle costs for equipment they have produced.

GE’s SmartSignal, for instance, is one such predictive analytics tool that employs digital twins to allow users from mining to manufacturing to diagnose, forecast, and prevent equipment failures ahead of time.^{14 ,15, 16}

Kieran Kelly, ubloquity (see case study below), noted, "Preventative maintenance is evolving from reactive to proactive strategies using AI to gain an understanding of failure patterns and accurately identify conditions for changing components."