Skip to main content
Loading navigation
Back to challenges

LNG Emissions Monitoring Agent with Google ADK & Amazon Bedrock

Operating ADNOC's new LNG export project in the UAE demands strict tracking of methane leak telemetry and flare emissions. Build an autonomous agent using Google Agent Development Kit (ADK) connected to Amazon Bedrock models to analyze multisensory greenhouse gas reports and auto-trigger environmental mitigation protocols. Achieve 98% detection accuracy for regulatory emission breaches.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Develop an LNG emissions compliance agent using Google ADK hosted on Amazon Bedrock models.

How work is evaluated

Evaluation

Evaluates detection precision of methane sensor anomalies and regulatory alert dispatching.

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

Threshold Breach Verification

Ensures breach_detected is true whenever methane_ppm exceeds 50.

Binary check

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

Dimension 2

Methane Leak Precision

Accuracy of classifying emission breaches correctly. • target: 0.98 • 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

  • Construct multi-step agent actions with Google ADK framework

  • Route inference calls to Amazon Bedrock Claude/Titan endpoints

  • Parse IoT gas leak sensor streams and satellite flare detection feeds

  • Issue automated UAE Ministry of Climate Change compliance alerts

Resources

Dataset notes

Telemetry datasets from UAE LNG export terminals including methane ppm, flare stack temperatures, and wind vector feeds.

Challenge setup

Evaluates detection precision of methane sensor anomalies and regulatory alert dispatching.

Challenge input

JSON object with sensor_id, methane_ppm, wind_speed_kmh, flare_temp_c

Google ADK

Assigned framework for agent structure.

Amazon Bedrock

Cloud foundation model execution infrastructure.

Evaluated output

JSON object with breach_detected, severity_level, mandatory_report_required

Evaluation checks
  • Ensures breach_detected is true whenever methane_ppm exceeds 50.
  • Accuracy of classifying emission breaches correctly. • target: 0.98 • range: 0-1
Configured evaluation evidence
  • Benchmark: EcoAgentBench
  • Methane Leak Precision target: 0.98
  • 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
  • Amazon Bedrock · Required
  • Google · Optional
  • Google ADK · Optional

Evaluation contract

  • Threshold Breach Verification · Weight 1
  • Methane Leak Precision · Weight 1

Metrics

  • Methane Leak Precision · Target 0.98

Benchmarks

  • EcoAgentBench
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with sensor_id, methane_ppm, wind_speed_kmh, flare_temp_c

    Output: JSON object with breach_detected, severity_level, mandatory_report_required

    Execution harness: Python

Evaluation tools

  • Google ADK · Agent Framework
  • Amazon Bedrock · LLM Inference Runtime

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about LNG Emissions Monitoring Agent with Google ADK & Amazon Bedrock