Skip to main content
Loading navigation
Back to challenges

Phosphate Rail Logistics Orchestrator with AutoGen

Saudi Arabia awarded an estimated $1bn phosphate rail track doubling deal. Build a multi-agent conversational system using Microsoft AutoGen where Logistics Agents, Track Operators, and Freight Schedulers negotiate train dispatching schedules. Increase throughput efficiency by 12% without track collision risks.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Implement conversational multi-agent collaboration with Microsoft AutoGen to optimize freight train schedules on Saudi Arabia's doubled phosphate rail corridor.

How work is evaluated

Evaluation

Evaluates freight train schedule output generated through AutoGen negotiation for safety and throughput.

Datasets

Public datasets, plus the private datasets you have access to.

How submissions are scored

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

zero_collision_check

Verifies safe headway separation between consecutive trains

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

throughput_efficiency

Total metric tons transported per day on schedule • target: 85000 • range: 0-100000

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

  • Construct a group conversation in AutoGen with `AssistantAgent` and `UserProxyAgent` roles.

  • Formulate custom consensus rules for train headway and station pass loops.

  • Incorporate rail network constraints for heavy mineral transport.

  • Validate autonomous negotiation outcomes against physical safety thresholds.

Resources

Dataset notes

Train movement and track topology data for Saudi Arabia's North-South mining railway network.

Challenge setup

Evaluates freight train schedule output generated through AutoGen negotiation for safety and throughput.

Challenge input

JSON containing list of train demand, origin, destination, and speed profile

AutoGen

Enables multi-agent conversational negotiation.

Evaluated output

JSON schedule with departure times and siding reservations

Evaluation checks
  • Verifies safe headway separation between consecutive trains
  • Total metric tons transported per day on schedule • target: 85000 • range: 0-100000
Configured evaluation evidence
  • Benchmark: MultiAgentRailBench
  • Throughput Efficiency target: 85000
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • AutoGen · Required
  • RAI · Optional
  • Decipher AI · Optional

Evaluation contract

  • zero_collision_check · Weight 1
  • throughput_efficiency · Weight 1

Metrics

  • throughput_efficiency · Target 85000

Benchmarks

  • MultiAgentRailBench
  • robotics

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing list of train demand, origin, destination, and speed profile

    Output: JSON schedule with departure times and siding reservations

    Execution harness: Python

Evaluation tools

  • AutoGen · Agent System Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Phosphate Rail Logistics Orchestrator with AutoGen