Global Battery Manufacturer Transforms Supply Chain Forecasting
A global manufacturer of batteries and portable lighting products faced major disruptions to its shipping and supply chain during the COVID-19 pandemic. Previously, the company relied on manual, Excel-based processes that proved inadequate as prices spiked and lead times increased drastically. To address this, Protiviti partnered with the manufacturer to introduce advanced predictive analytics using Microsoft Azure, enabling rapid, accurate forecasting for 63 crucial shipping lanes. Data governance and cleansing methodologies were implemented, and hands-on training empowered the client's team to build, refine, and operationalize low-error predictive models. This data-driven transformation allowed the company to become highly agile in responding to market shifts, increase expense management accuracy, and become a pioneer in predictive analytics for the consumer goods shipping industry. The Azure-powered solution also laid the foundation for improved performance in related shipping and operations functions, permanently elevating the client's analytics capabilities.
- Industry
- Manufacturing
- Location
- Global
- Published
- September 2023
Reported outcomes
Strategic outcomes
Primary read
Use case focus
Showing 3 of 3
- 1Predictive Analytics for Ocean Freight Cost Forecasting
- 2Supply Chain Cost Optimization using ML
- 3Shipping Lane-specific Price Prediction
- Extreme volatility in ocean freight prices and lead times during COVID-19 pandemic.
- Manual Excel-based shipping cost analysis tools led to limited forecasting visibility and agility.
- Frequent stockouts and operational disruptions due to slow response times.
- Lack of reliable data governance and standardized definitions across shipping operations.
- The inability to accurately forecast cost trends and risks threatened competitiveness.
- Introduced Microsoft Azure Machine Learning for predictive analytics across 63 shipping lanes.
- Partnered with Protiviti for data cleansing, governance, and analytics training.
- Used Python in Azure for data ingestion, model building, and automation.
- Leveraged Microsoft AutoML for hyperparameter tuning and rapid test iteration.
- Standardized nomenclature and implemented robust data governance across shipping data.
- Improved ocean freight forecasting accuracy with lane-specific low-error ML models.
- Achieved greater agility in shipping and cost management decision-making.
- Reduced manual workload and increased analytics adoption within business teams.
- Enabled actionable insights for 30, 60, and 90-day forecasts, leading to optimized shipping operations.
- Established robust data governance supporting more reliable and scalable analytics programs.
Architecture
The solution ingested raw ocean freight and shipping data into Microsoft Azure. Python scripts running in Azure processed, cleansed, and standardized the data to ensure accuracy for every source-destination shipping lane. Azure Machine Learning was used to develop lane-specific predictive models, and Microsoft AutoML provided hyperparameter tuning and automated model selection. The output informed downstream operations through clean, standardized forecasts and was further leveraged to establish data governance practices throughout the organization.
Implementation partners1
Sources & evidence1
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Was this useful?