Fair Workload Distribution Engine
Python · Algorithms · Operations
Balanced output across ~700 units/day
The problem
On a real factory floor, the obvious way to assign work is also the worst one: hand each order to whoever’s free or fastest. Do that and your best operators get buried, everyone else sits idle, and output swings wildly from day to day. On the Sirwaal line alone — ~700 units a day — that imbalance quietly costs capacity, quality, and morale.
What I built
A fair-distribution engine inside the Mawan ERP. Instead of assigning by speed, it models each team’s current load and remaining capacity, then places every new production order where it keeps the whole floor balanced. It accounts for work already in progress, so no one gets handed a queue they can’t clear. The results stream into live Power BI dashboards, so supervisors watch throughput and balance as the shift happens — not reconstruct it afterward.
The result
A floor that’s fairer and more predictable: steady output across teams handling ~700 units a day, less burnout on the top performers, and full visibility into every line.