Skip to main content
Loading navigation
Back to challenges

Refinery Incident Repair Orchestrator with LangChain & LangGraph

Refinery maintenance teams responding to physical damage events require automated task graph generation for emergency repairs. Build a stateful incident remediation graph using LangChain and LangGraph to schedule contractor tasks and track equipment replacements.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Construct a stateful, multi-node repair planning graph using LangChain and LangGraph for industrial facility recovery.

How work is evaluated

Evaluation

Evaluates graph transition logic, critical path duration calculations, and state persistence.

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

stage_progression_check

Verifies that state finishes in 'PROCUREMENT_AND_INSTALL' or 'COMPLETED'.

Binary check

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

Dimension 2

downtime_calculation_precision

Accuracy of total estimated downtime relative to lead times and repair windows. • target: 0.95 • range: 0-1

Binary check

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

Learning goals

  • Design a cyclic, stateful workflow using LangGraph for multi-stage repair planning

  • Incorporate LangChain tool interfaces to query contractor resource availability

  • Implement checkpointing and state memory for long-running infrastructure maintenance

  • Evaluate schedule critical path impact based on replacement lead times

Resources

Dataset notes

Damage assessment logs, replacement valve/pipe inventory records, and contractor mobilisation timetables for Kuwait refineries.

Challenge setup

Evaluates graph transition logic, critical path duration calculations, and state persistence.

Challenge input

JSON with unit_id, damage_severity, and lead_time_days

LangChain

Standard components for LLM prompt orchestration.

LangGraph

Stateful graph orchestration engine for multi-node agent logic.

Evaluated output

JSON with total_downtime_days, active_stage, and contractor_assigned

Evaluation checks
  • Verifies that state finishes in 'PROCUREMENT_AND_INSTALL' or 'COMPLETED'.
  • Accuracy of total estimated downtime relative to lead times and repair windows. • target: 0.95 • range: 0-1
Configured evaluation evidence
  • Benchmark: Industrial-Graph-Ops
  • Downtime Calculation Precision target: 0.95
  • 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
  • Langchain · Required
  • LangChain · Optional
  • Merge · Optional
Orchestration
  • Langchain · Required
  • LangChain · Optional

Evaluation contract

  • stage_progression_check · Weight 1
  • downtime_calculation_precision · Weight 1

Metrics

  • downtime_calculation_precision · Target 0.95

Benchmarks

  • Industrial-Graph-Ops
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with unit_id, damage_severity, and lead_time_days

    Output: JSON with total_downtime_days, active_stage, and contractor_assigned

    Execution harness: Python

Evaluation tools

  • LangChain · llm_tooling
  • LangGraph · agent_framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Refinery Incident Repair Orchestrator with LangChain & LangGraph