Automated work order routing on shop floor improved production efficiency by 35%

Sharing our success story implementing automated work order routing in our manufacturing facility. We’re a mid-sized electronics manufacturer running Odoo 15 Manufacturing with 4 production lines and about 80 workstations.

Previously, production supervisors manually assigned work orders to workstations each morning based on priority, capacity, and worker skills. This process took 45-60 minutes daily and often resulted in suboptimal assignments because supervisors couldn’t consider all variables simultaneously.

We implemented an automated routing workflow using Odoo’s Workflow Engine that considers workstation capacity, worker certifications, current queue depth, and order priority. The system automatically assigns incoming work orders to the optimal workstation and notifies workers via the shop floor tablet interface.

Results after 3 months:

  • Average lead time reduced from 4.2 days to 2.7 days
  • Workstation utilization increased from 68% to 87%
  • Setup time between jobs decreased 22% due to better skill matching
  • Supervisor time freed up for quality improvement initiatives
  • On-time delivery improved from 83% to 96%

The implementation took about 6 weeks including workflow design, testing, and operator training. Total investment was roughly $15K including consultant time, and we achieved payback in under 4 months through efficiency gains.

Impressive results! How did you handle the worker certification tracking? Is that built into Odoo 15 or did you need custom development? We have a similar challenge with skilled trades requiring specific certifications for certain operations.

The 35% efficiency improvement is remarkable. Can you share more details about how you measured baseline performance and tracked improvements? Also curious about employee acceptance - did you face any resistance to the automated assignments?

This is exactly the level of detail we needed. The phased approach and continuous refinement based on override data is brilliant. We’re going to use your implementation as a model for our client project. Thanks for sharing so openly!

Great questions - let me provide more detail on the implementation:

Worker Certifications: We used Odoo’s built-in Skills module (available in Enterprise 15+) to track certifications. Each operation in our routing is tagged with required skills, and workers have skill profiles with proficiency levels and expiration dates. The routing algorithm only assigns work orders to qualified workers. We did add a custom field to track certification renewal dates and built an automated reminder workflow for expiring certs.

Measurement & Acceptance: We established baseline metrics during a 4-week pre-implementation period tracking: work order completion time, queue time at each workstation, and setup time between jobs. Post-implementation, the same metrics are automatically captured in Odoo’s MRP analytics. For employee acceptance, we involved floor supervisors and lead operators in the workflow design phase which built buy-in. The key selling point was that automation handles routine assignments, freeing supervisors for exception handling and continuous improvement. Workers actually love it because assignments are fairer and consider their skill preferences.

Override Capability: Absolutely critical. Supervisors can override any assignment through the MRP dashboard, and we require them to select a reason code (rush order, equipment issue, skill development, etc.). These overrides are logged and reviewed monthly. We use that data to refine the routing rules - for example, we discovered we were underweighting setup time in the initial algorithm and adjusted based on override patterns.

Rush Orders: The routing algorithm includes a priority weighting factor. Rush orders (tagged with high priority in sales order) get a 3x multiplier in the assignment calculation, which typically results in them being assigned to the next available qualified workstation. Supervisors can also manually trigger re-optimization of the current queue when a hot order comes in, which may reassign lower-priority jobs to accommodate the rush.

Kanban Integration: Yes! This was phase 2 of our implementation. We integrated the automated routing with kanban triggers so that work orders are automatically generated and routed when inventory hits reorder points. This created a true pull system. The combination reduced our WIP by 40% while maintaining the same throughput.

Technical Architecture: The routing engine runs as a scheduled action every 15 minutes during production hours, evaluating new work orders and current workstation status. It uses a weighted scoring algorithm considering: worker skill match (30%), current queue depth (25%), setup time from previous job (20%), workstation preference for the operation (15%), and worker availability (10%). The weights are configurable and we’re still fine-tuning based on actual performance data.

The biggest lesson learned: start simple with the routing logic and add complexity gradually based on real-world feedback. Our initial algorithm was too complex and made some counterintuitive assignments. We simplified it, ran for a month, then added refinements based on supervisor overrides and performance data.

Happy to share more specifics if anyone is considering a similar implementation.

What happens when the automated system makes a clearly wrong assignment? Do supervisors have override capability, and if so, does that data feed back into the routing algorithm for continuous improvement?