Supply-chain AI and automation advance faster than data integration, GS1 US survey finds

GS1 US published survey findings on data integration and AI or automation in supply chains.

A study published November 3 by GS1 US found that nearly 60% of surveyed organizations had either automated most supply-chain processes or implemented AI or machine learning. Only 38% reported a centralized system that collects and integrates data in real time. The gap points to a challenge for teams trying to move automation beyond pilots. A model or rule can make a suggestion quickly, while the inventory, order or supplier data behind it still arrives late or requires manual reconciliation. [GS1 US research release](https://www.gs1us.org/industries-and-insights/media-center/press-releases/ai-and-automation-investments-outpacing-supply-chain-data-readiness)

What the researchers measured

Redpoint surveyed 500 supply-chain professionals for GS1 US across industries including grocery, healthcare, foodservice, apparel and consumer packaged goods. Thirty-two percent reported manually bringing together information from multiple sources. Forty-one percent said they could pilot a new technology but had trouble expanding its use. Respondents cited fragmented data and funding constraints at nearly the same rate. The headline adoption figure combines advanced automation with AI and machine learning, so it cannot establish how many organizations use AI alone. [GS1 US research release](https://www.gs1us.org/industries-and-insights/media-center/press-releases/ai-and-automation-investments-outpacing-supply-chain-data-readiness)

The researchers grouped organizations by measures spanning integration, governance, collaboration, automation and use of standards. Among organizations the researchers placed in the highest maturity group, 81% reported centralized, real-time data integration. That comparison describes the sample. Because integration was among the inputs to the grouping, a higher rate in that tier should not be read as independent proof that a particular data architecture drives better business results.

Where the gap shows up in practice

Consider a replenishment alert. A planning tool might recommend placing a new order because it sees low stock at one warehouse. If the order system already has an inbound shipment that the planning tool cannot see, the alert can create excess inventory. Fixing that gap involves matching records across systems and knowing when the underlying quantities were last updated. A better prediction model alone cannot supply a missing transaction.

A rollout can therefore track a narrow set of operational measures alongside model accuracy. Examples include the age of inventory records used in recommendations, how often staff correct mismatched product identifiers and how many proposed actions are reversed because the underlying data changed. Those measures can help explain why an experiment performs well at one site and stalls when expanded across teams or trading partners. The survey documents the prevalence of such integration challenges among respondents. It leaves the return on individual projects to be measured in their own operating conditions. [GS1 US research release](https://www.gs1us.org/industries-and-insights/media-center/press-releases/ai-and-automation-investments-outpacing-supply-chain-data-readiness)