Logistics Company

Route optimisation with AWS SageMaker

Starting point

  • The Faber Group BV offers pallet and pooling services for a variety of industries
  • Customers are provided with tailored solutions for their different pallet requirements
  • The demand for shorter delivery times requires optimized routes
  • Rising fuel prices and environmental regulations increase the pressure on the company
  • The complexity of supply chains requires intelligent solutions

Role in the project

  • Creation of a detailed requirements analysis
  • Provision of a business solution design
  • Technical design for the integration of SageMaker
  • Development of the product-specific and customized customer solution
  • Through the integration of Amazon SageMaker, a machine learning-based model was quickly and easily implemented

Features/Project outcome

  • Development of a custom SageMaker model for machine learning
  • Implementation of a Python algorithm for many-to-many route optimization
  • Integration of the OpenRouteService API for geodata
  • Saving the optimal routes and matching order routes with historical routes
  • Limiting the results through additional inputs such as start and destination

Customer benefits

  • Significant cost savings in the double-digit million range
  • Optimisation of vehicle utilization and reduction of emissions
  • Better planning and higher flexibility
  • Increased customer satisfaction due to faster delivery times
  • Expansion of details to narrow down search queries:
  • Subsections < 200 km; total distance < 3,200 km
  • Optimisation over n>1 trucks
  • Customer number instead of invoice number for the search algorithm
Lars Ritter | Senior Manager