Operations pilot uses AI to monitor and classify activity across industrial conveyor belt systems in factories and recycling facilities.

Through operational data, the AI detects patterns and categorizes facility activity into different operational states:

• Normal production activity

• Planned breaks

• Maintenance periods

• Downtime events

• Unexpected interruptions

This information is used to generate operational reports and help teams understand productivity, performance, and inefficiencies across facilities.

However, the AI depends on accurate manual schedule input from operations managers.

Shift schedules, breaks, maintenance windows, and operational changes need to be manually configured so the system can correctly understand why a facility appears inactive.

Without this information:

• planned breaks may be interpreted as downtime

• reports become inaccurate

• operational insights lose trust and reliability

• teams make decisions using incorrect data

The challenge was to design an experience that allows managers to create, manage and modify operational schedules efficiently while supporting both long-term planning and real-time changes.