AI / MLOEE & production

AI-based shift planning

Automotive Germany

An ML model replaces manual shift planning — planning effort halved, overtime significantly reduced.

A mid-sized automotive supplier with three production lines was spending several hours every day on manual shift planning. Holiday peaks, qualification profiles and machine capacities all had to be coordinated by hand.

RALUS developed an ML-based planning system that combines historical shift data, qualification profiles and machine status. Every day the model proposes optimised shift plans — the planner validates them in minutes instead of hours.

After a six-week rollout, the system was fully integrated into daily operations. Planning effort fell by 50%, and overtime caused by poor staffing by 31%.

Results

−50%
Planning effort
−31%
Overtime
6 wks
Rollout

Technologies we use

PythonFastAPIReactPostgreSQLDocker
#Shift planning#ML#Resource optimisation#Python