[ANN] JuLS - Julia Local Search solver from Amazon Science

Package Announcements
,
1

Behind Amazon’s logistics operations lie some of the most complex combinatorial optimization challenges in industry. Curious how we solve it? We built a generic solver from scratch, and today, it’s open-source. :rocket:

Meet JuLS (Julia Local Search), Amazon Science’s new open-source solver: Developed by our team of scientists, JuLS is fast, extensible, and welcomes community contributions. It combines Constraint-Based Local Search and Constraint Programming, and it’s already deployed in Amazon’s production environment to tackle middle-mile routing and scheduling.

JuLS performs on par with OR-Tools on standard tasks, but its real strength is handling black-box objectives and constraints, making it a powerful solution for real-world optimization scenarios that often require interfacing with proprietary or external systems. More than a solver, JuLS is a playground for researchers and practitioners tackling combinatorial optimization problems.

:backhand_index_pointing_right: Check it out & drop us a :star:: GitHub - amazon-science/JuLS: JuLS is a Julia Local Search solver that combines Constraint Based Local Search (CBLS) and Constraint Programming (CP)

18 Likes
2

Nice package! Happy to hear Amazon is using Julia somewhere!

Looking at the Project.toml file, it looks like:

  1. you have developing/testing packages checked in as well (Revise, Test）
  2. you don’t have [compat] set
2 Likes
3

This is really exciting and fun to check out. Can’t wait to see if there are plans to make the model construction (DAG generation) more user friendly for new optimization problems, either by some package specific DSL or even integration with JuMP!

2 Likes