From forecasting future demand for products to identifying supply chain bottlenecks before they happen, we look at how AI could help manufacturers optimize outbound logistics.
AI is increasingly helping to automate core logistics processes.
Already, manufacturers can use machine learning to analyze news and reviews to assess distributor risk, AI-driven marketplaces can match manufacturers with third-party logistics providers, and machine learning can help evaluate distributor contracts. (Read about how similar approaches are being used for sourcing in part 2 of this series.)
FREE DOWNLOAD: THE COMPLETE AI 100 LIST
Get an excel file with the entire AI 100 list including each company’s total funding, focus area, and more.
But other logistics challenges for manufacturers persist, including gaining real-time feedback for product demand, monitoring the transport of manufactured goods, and harvesting data directly from their fleets. In response, startups are developing AI-driven services and are partnering with CPG manufacturers and industrial OEMs to take on these challenges.
In part 5 of our series on AI in manufacturing, we look at AI use cases in forecasting demand, tracking manufacturing activity using geospatial intelligence, and analyzing real-time telematics data.
- AI-driven demand forecasting and retail feedback services are helping manufacturers to calibrate their production in real time. Startups such as o9 Solutions use AI-generated “digital twins” and scenario planning to inform production and distribution decisions.
- Using AI software and satellite imagery, geospatial intelligence tools can track activity and abnormalities in factories and monitor the flow of raw materials in supply chains. Startups such as Descartes Labs use satellite imagery to trace materials and goods along the manufacturing supply chain.
- AI-driven fleet telematics provide manufacturers with real-time awareness of the state of their distribution channels. Startups such as Uptake are combining AI software with sensor data to identify vulnerable machine components and prevent fleet downtime.
Why outbound logistics?
While improved connectivity has closed the feedback gap between what retailers are selling and what manufacturers are producing, the conventional processes for relaying information can cause delays that make it difficult for manufacturers to alter production quickly.
Additionally, traditional demand forecasting methods — which rely on historical sales data and existing inventory — do not adequately incorporate the impact of real-time events or unusual fluctuations, as evidenced during the early months of the Covid-19 lockdowns.
Further, while improvements in location technology have enabled manufacturers to better monitor their supply chains, many players are now seeking more granular data points to help predict supply constraints and distribution bottlenecks.
Demand forecasting for outbound logistics
Demand forecasting is a key component of manufacturing procurement and distribution. Supply chain data, ranging from product volume and price data to more granular information about the state of warehouses and stores, can improve outbound logistics decision-making.
Machine learning (ML) enables manufacturers to more easily identify relationships between sales and SKU-level data, store-level data, large-scale economic events, and alternative data sources.
Attention for ML-based demand forecasting has surged in recent years and the market is anticipated to reach $3B by 2023, according to CB Insights’ Industry Analyst Consensus.
o9 Solutions, a Dallas-based unicorn, offers AI-driven supply chain management services for manufacturers including Nestle, Kraft Heinz, Pirelli, and General Electric. The company’s platform uses AI to create a digital twin — a virtual replica — of a manufacturer’s supply chain, which can then help inform production and distribution decisions, enable teams to collaborate using scenario planning simulations, and identify production constraints.