Back to evaluations
Public evaluation
aeration_control_eval
Evaluates the agent's ability to lower energy usage while maintaining dissolved oxygen (DO) above 2.0 mg/L across simulated aeration scenarios.
Evaluation type
task based
Challenge
Wastewater Aeration Optimization Agent with LangChain and OpenAI o3
Difficulty
Intermediate
Rigor
Unspecified
Evaluation overview
How the linked challenge is judged: tasks, benchmarks, and criteria count.
Tasks
1
Benchmarks
0
Criteria
0
Task templates
Inputs and expected outputs.
Task 1
aeration_control_eval
Tests agent responses to dynamic influx of biochemical oxygen demand (BOD) and DO telemetry.
Input format
JSON object with time-series DO values, inflow rate, and current blower frequency.
Output format
JSON object with target_blower_frequency_hz and reasoning_summary.