AI System for Identifying Capacity Exceptions in Production Plans

The problem

When orders, routings, machines, moulds and shift plans are held in separate sources, their fields must be matched by hand before resource conflicts and draft alternatives can be prepared.

The production planner must compare every suggested resource conflict and draft alternative with its source record; missing or conflicting evidence must not replace the final decision.

If an alternative machine that is suitable only for certain revisions is not set aside for review, an unworkable plan could be released; the record must not proceed without a human decision.

What we set out to improve

• Time taken for draft resource conflicts, alternatives and their supporting evidence to reach the production planner for review
• How records set aside for human review because an alternative machine is suitable only for certain revisions are resolved
• Number of suggestions corrected after identifying a risk that an unworkable plan could be released

How the system works

1. 01 Retrieve orders, routings, machines, moulds and shift plans from permitted sources using the record identifier
2. 02 Link source, date and version information to the shared work record
3. 03 Validate fields and units
4. 04 Prepare the resource conflicts and draft alternatives with their supporting evidence
5. 05 Check for an exception: an alternative machine is suitable only for certain revisions
6. 06 Have the production planner accept, revise or reject the suggestion
7. 07 Record the approved decision and its source in the processing log

Methods we used

• A source-identified common data schema for orders, routings, machines, moulds and shift plans
• A decision structure that separates deterministic rules from AI suggestions
• A review screen showing resource conflicts and draft alternatives alongside the underlying source evidence
• A mandatory human review queue when an alternative machine is suitable only for certain revisions
• An event identifier that prevents duplicate processing

Where people stay involved

Data and security

Only the fields required for this decision should be processed from orders, routings, machines, moulds and shift plans. Access to the source systems must be authorised separately from the production planner role, and personal data, commercial information and technical secrets must not be passed into the model context unnecessarily. Reading, suggestions and human decisions must be logged separately.

Who this suits

Frequently asked questions