P&G Cuts Shift Planning from 8 Hours to 20 Minutes: The Case for Integrated Data

On 6 November 2026, IIoT World published an article based on a CxO series video interview with Keith Moore, CEO of AutoScheduler.AI. It describes how Procter & Gamble reduced daily workforce planning in its warehouses from about 8 hours per shift to 20 minutes. One caveat up front: the source is the vendor itself, and the results have not been independently verified.

The problem: a plan that lasts ten minutes

At P&G's production campuses, preparing a single shift plan took about eight hours. Often the plan fell apart within ten minutes of the shift starting, because of equipment breakdowns, absences or late trucks. Shuttle movements between buildings must be precisely synchronized, so every delay forced a re-plan.

According to Moore, the root cause is siloed legacy systems. A labor management system and a separate slotting solution rarely talk to each other, even though labor efficiency depends on where the goods are and when the trucks arrive.

The solution: a semantic layer and the local 20%

AutoScheduler.AI spent six years building a semantic layer that encodes how a warehouse should ideally operate and links WMS and ERP data with equipment capacity and dock schedules. This layer automates roughly 80% of decisions, which are common to all warehouses, and it is what brought planning down from 8 hours to 20 minutes.

The remaining 20% is specific to each plant: for example, a sorter with particular loading rules or a cross-dock flow that is active only on certain days. This is where department supervisors, with no programming background, come in: using an AI-based no-code tool, they have built almost 500 custom apps. The traditional route would be an IT development cycle of about six months.

A concrete example

When a load is short of product, one of these apps checks SAP/ERP data on upcoming shipments to that customer and suggests substitutions to protect the fill rate. It replaced a manual task performed by two people and, according to the account, the supervisor who came up with it saved the plant $150,000 within two days of going live.

Trust comes from transparency

The AI records its own reasoning in plain language. An experienced supervisor can reconstruct why a recommendation was made, challenge it and adjust its parameters.

A recommendation that can be explained and corrected is more likely to be used on the shop floor.

The advice: map decisions, don't clean data forever

Moore advises against getting trapped in endless data cleanup. It is better to start from a map of decisions:

For each area, ask:

• what is being decided;
• why it is decided that way;
• which data feed the decision.

Then choose the highest-value areas to automate, leaving fragile spreadsheets behind.

What this means for MES and monitoring

The case concerns warehousing and logistics, not production lines directly, and the available information does not indicate costs, implementation times or the number of sites involved. The lessons, however, are transferable:

• Integrating data across systems removes the bottleneck more effectively than adding point tools.
• Standardize common decisions and leave local exceptions to the people who work on the floor.
• Make the logic transparent to earn the trust of your teams.
• Map decisions before data: the same approach applies to MES and production monitoring projects.

Before evaluating any technology, it is therefore worth asking which decisions are made by hand today, with which data and which spreadsheets: it is the most concrete starting point for a digitalization project.