Automating on-demand logistics operations for scale, customization, and iteration is more achievable than you might think. Learn how to build, test, and deploy models for demand forecasting, shift scheduling, and route creation using OR-Tools (plus LAD regression), HiGHS (with MIP), and Nextroute (with ALNS).
Ryan O’Neil, CTO at Nextmv, will walk through working code for each model, explain how they work together, cover testing techniques for de-risking production rollout, and explore opportunities for rapid iteration and continuous improvement. By the end of the talk, we have assembled everything needed for a working logistics service.
- How distinct logistics models can be created, linked, and managed from a single platform
- Optimization is a diverse world, but that doesn’t necessitate complex infrastructure
- How to use Google OR-Tools and Python to forecast and link the output to a HiGHS (MIP) scheduling model and a heuristics-based routing model