01 / The problem

Every plant has rules that never make it into a system. Line 3 doesn't run that resin on Mondays. This customer always ships first. Those two products never run back to back. In small plants, those rules live in a planner's head.

Scheduling software can only plan around what it's told. Getting all of that into a system takes time small manufacturers don't have, so most keep scheduling in spreadsheets.

02 / What we're working on

We're researching scheduling that gets better through normal use. When a planner moves a run to a different line or time, that correction says something about how the plant really works. Our goal is software that learns from those corrections and asks a quick question when it isn't sure, instead of making people configure everything up front.

03 / Where we start

We focus on process manufacturers with 2 to 15 production lines, starting with plastics extrusion, where product changeovers make the order of runs matter.