Skip to main content
Loading navigation
Back to challenges

Waste-to-Energy Dispatch Agent with OpenAI Agents SDK

Municipal waste facility operators need automated load balancing to maximize energy output while remaining compliant with environmental emissions thresholds. Build an autonomous waste-to-energy dispatch agent using the OpenAI Agents SDK routed through Vercel AI Gateway to maintain combustion zone stability.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Design an autonomous dispatch agent using the OpenAI Agents SDK and tool functions to regulate waste feedstock feed rates.

How work is evaluated

Evaluation

Evaluates agent capability to adjust waste feeding rates based on caloric telemetry without exceeding emissions caps.

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

temperature_stability_check

Ensure combustion temperature remains between 850C and 1050C.

Binary check

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

Dimension 2

feed_rate_accuracy

Absolute error between calculated optimal feed rate and agent recommendation. • 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

  • Implement autonomous tool invocation using OpenAI Agents SDK

  • Route agent model traffic through Vercel AI Gateway for resilience and observability

  • Parse real-time waste sensor telemetry and calculate optimal furnace feed rates

  • Enforce environmental compliance boundaries within dynamic tool function logic

Resources

Dataset notes

Simulated telemetry dataset containing waste caloric values, moisture content, furnace temperatures, and target emissions metrics for waste-to-energy plants.

Challenge setup

Evaluates agent capability to adjust waste feeding rates based on caloric telemetry without exceeding emissions caps.

Challenge input

JSON object with caloric_value_kj_kg, moisture_content_pct, and current_temp_c

OpenAI Agents SDK

Core framework required for agent orchestrations.

Vercel AI Gateway

Provides proxy routing and monitoring for LLM API calls.

Evaluated output

JSON object with recommended_feed_rate_tph and airflow_ratio

Evaluation checks
  • Ensure combustion temperature remains between 850C and 1050C.
  • Absolute error between calculated optimal feed rate and agent recommendation. • target: 0.95 • range: 0-1
Configured evaluation evidence
  • Benchmark: EnergyOps-Eval-v1
  • Feed Rate Accuracy target: 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
  • OpenAI · Required
  • VAPI · Optional
Policy Serving
  • Vercel AI Gateway · Optional
Orchestration
  • Vercel AI Gateway · Optional

Evaluation contract

  • temperature_stability_check · Weight 1
  • feed_rate_accuracy · Weight 1

Metrics

  • feed_rate_accuracy · Target 0.95

Benchmarks

  • EnergyOps-Eval-v1
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with caloric_value_kj_kg, moisture_content_pct, and current_temp_c

    Output: JSON object with recommended_feed_rate_tph and airflow_ratio

    Execution harness: Python

Evaluation tools

  • OpenAI Agents SDK · agent_framework
  • Vercel AI Gateway · llm_gateway

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Waste-to-Energy Dispatch Agent with OpenAI Agents SDK