TOPIC
Estimating Arrival Time in presence of Uncertainty using Machine Learning: Case of Container Tracking for Bosch PT Supply Chain Operations
Miodrag Ivanović & Najdan Vuković
BI Tech Lead & AI&ML Tech Lead, Senior Machine Learning Engineer at Bosch
About Miodrag & Najdan
Miodrag Ivanović is a Business Intelligence Tech Lead, currently working on project Supply Chain Analytics for Bosch Power Tools division. Miodrag has more than 16 years of experience with DWH technologies, mainly on Microsoft BI stack. Also, Miodrag is MCT (Microsoft Certified Trainer) for Microsoft Data platform.
Najdan is an Artificial Intelligence and Machine Learning Tech Lead in Bosch, currently works on forecasting of thousands of timeseries based on Machine Learning. Najdan has PhD in Machine Learning, 10+ years of research experience in science and academia, and 7+ years of experience in implementation and development of Machine Learning models for business applications.
Talk Description
As a global company, BOSCH Power Tools (BOSCH PT) needs to transport goods and products, meet customers needs and maintain profit margin. Majority of its products are transported by ships, however, due to various challenges carriers have problems meeting deadlines. This induces heavy burden on global operations because each ship being late results in increased costs. To mitigate this, we developed machine learning model provides Estimated Arrival Time (ETA) for each ship carrying BOSCH PT’s containers, which significantly helps BOSCH PT to adjust its daily operations in accordance with sudden and unpredictable changes. Currently, our solution runs on-premise but we are looking forward to extend it to cloud infrastructure.