Challenge

OpenAI Agents SDK PLC Telemetry Anomaly Detector with Qwen 3

Wastewater aeration retrofits require reliable telemetry monitoring to prevent dissolved oxygen control failures. Build an agentic monitoring agent using the OpenAI Agents SDK and Qwen 3 to detect telemetry anomalies across small PLC retrofits. The solution must reduce false alarms by 35% while flagging real aeration faults within 2 seconds.

EngineeringHosted by Vera
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Build an autonomous PLC telemetry anomaly detector using OpenAI Agents SDK tool calls and Qwen 3 model execution.

Delivery guide

How work is evaluated

Evaluation

Evaluates agent anomaly detection precision, latency, and function calling validity against simulated PLC telemetry logs.

Datasets

Shared data for this challenge

Review public datasets and any private uploads tied to your build.

Loading datasets...
Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

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

schema_conformance

Verifies output structure matches exact telemetry evaluation output schema

Binary check

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

Dimension 2

anomaly_precision

Precision score of anomaly classification • target: 0.92 • range: 0-1

Binary check

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

Dimension 3

latency_ms

Average response latency per packet in milliseconds • target: 300 • range: 0-2000

Binary check

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

Learning goals

What you should walk away with

  • Implement real-time PLC telemetry stream ingestion with OpenAI Agents SDK

  • Configure Qwen 3 function calling to identify dissolved oxygen anomalies

  • Develop root-cause analysis logic for wastewater aeration blower controls

  • Validate automated alert triage to reduce false positive rates below 5%

Resources and assets

Reference links and supporting material

Dataset notes

Simulated 24-hour time-series telemetry dataset of a wastewater treatment plant's aeration basin equipped with small PLC controllers.

How this agent runs

Evaluates agent anomaly detection precision, latency, and function calling validity against simulated PLC telemetry logs.

Challenge input

JSON lines with timestamp, dissolved_oxygen, air_flow_rate, blower_rpm, valve_position

OpenAI Agents SDK

Required framework for building autonomous function-calling agents.

Qwen 3

Open-weights model assigned for domain reasoning and function calling.

Evaluated output

JSON object with anomaly_detected (bool), root_cause (string), confidence (float), latency_ms (int)

Checks for
  • Verifies output structure matches exact telemetry evaluation output schema
  • Precision score of anomaly classification • target: 0.92 • range: 0-1
  • Average response latency per packet in milliseconds • target: 300 • range: 0-2000
Proof of success
  • Benchmark: Industrial Telemetry Benchmark v1
  • Anomaly Precision target: 0.92
  • 1 public reference case
Runtime evidence
  • Python execution harness
  • Python sandbox
View technical recipe

Configured tools

Action Space
  • OpenAI · Required
  • Alibaba Cloud (Qwen) · Optional
Policy Serving
  • Qwen 3 · Optional

Evaluation contract

  • schema_conformance · Weight 1
  • anomaly_precision · Weight 1
  • latency_ms · Weight 1

Recipe state

This is a preview. The configuration can change before the evaluation recipe is locked.

Frequently Asked Questions about OpenAI Agents SDK PLC Telemetry Anomaly Detector with Qwen 3