United StatesManufacturing2016 – ongoingJabil
Machine learning predicts manufacturing process failures
Jabil reported 17% scrap and rework savings after deploying factory failure prediction
Problem
Manufacturing faults detected late in assembly created scrap and rework. Jabil wanted earlier warning of processes likely to slow down or fail.
Approach
Jabil built a predictive system using Microsoft Azure Machine Learning. It analyzed machine data across manufacturing steps to identify failures earlier. The platform operated at sites in Penang, Malaysia, and Guadalajara, Mexico.
Results
Takeaway
The reported savings followed deployment at 2 sites. Worldwide deployment remained a plan at the time.
Categories
- Size
- Large enterprise
- Industry
- Manufacturing
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Partner co-development
- Evidence
- Company's own claim
Sources
Source summaries
Microsoft described the deployed prediction platform and its manufacturing data inputs. Jabil's quality executive supplied the prediction accuracy and scrap and rework savings figures.
Full details are in the original sources.