CLLIB - A Benchmark Library for Two-Tier City Logistics

The CLLIB is a benchmark library of instance data of planning problems in two-tier city logistics for academic use. It contains 3,564 instances different, e.g., in the underlying geographical region, size and transportation means. The underlying urban regions are:

District Oststadt of Karlsruhe, Germany (about 21,000 inhabitants and 5 km^2),

Ingolstadt, Germany (about 141,000 inhabitants and 133 km^2),

Munich, Germany (about 1.5 million inhabitants and 311 km^2).

The instances involve between 10 and 500 demands, road- as well as rail-based transport and temporal as well as resource constraints like demand time windows (TWs), maximum waiting times and capacities of vehicles and locations.

The CLLIB is generated using the CL Instance Generation Framework (coded in Python, publicly available via Zenodo). 

Sub-libraries are provided filtering the instances

by the urban regions (CLLIB-KA, CLLIB-ING, CLLIB-MUN),

by the number of demands (CLLIB-Small, CLLIB-Medium, CLLIB-Large), and

by the demand types involved (CLLIB-OnlyIn, CLLIB-InOut, CLLIB-AllDemTypes). 

See details in manual.

A library extension CLLIB-Analytics with additional 93,150 instances is provided to enable a proper evaluation of effects of single instance characteristics. See details in manual.