Updated

Updated · Google Research · Sep 18

Google Launches MilleMiglia C++ Generator for 450-Mile Middle-Mile Logistics Benchmarks

Updated

Updated · Google Research · Sep 18

Google Launches MilleMiglia C++ Generator for 450-Mile Middle-Mile Logistics Benchmarks

1 articles · Updated · Google Research · Sep 18

Google released MilleMiglia as an open-source C++ instance generator that creates realistic, privacy-preserving benchmarks for middle-mile delivery networks—the long-haul segment between distribution centers.
The tool targets a key research gap: middle-mile optimization has lacked public, high-quality data even though that segment covers the longest distances and a large share of logistics costs.
Unlike first- and last-mile vehicle-routing problems, middle-mile shipments can switch trucks across multi-day networks, forcing synchronization around fixed schedules, hub throughput limits and missed connections.
MilleMiglia uses Protocol Buffers to package compact benchmark instances for solvers in different languages and can also generate large datasets for machine-learning training.
Google said the project is a first step toward a standardized benchmark suite, with a specialized solver, API and possible research challenge planned through its collaboration with UniBrescia and ENPC Paris.

Sources

Will MilleMiglia's open-source approach finally break the long-standing data monopoly held by massive enterprise supply chain operators?

Could Google's new synthetic data generator accidentally train logistics algorithms for a predictable world that does not actually exist?

How can researchers guarantee this privacy-preserving logistics data cannot be reverse-engineered to expose highly confidential corporate networks?

1 total