BESS Health Monitoring Agent
As Battery Energy Storage Systems (BESS) mature, moving from commissioning to operational longevity requires sophisticated predictive maintenance. This challenge tasks you with building a Mastra AI-powered agentic workflow that analyzes high-frequency telemetry data (voltage, temperature, state of charge) to detect early-stage cell degradation and thermal anomalies. You will integrate CodeCarbon to measure the environmental impact of your AI inference and training cycles, ensuring that the 'Green AI' solution does not consume excessive energy while monitoring renewable assets. This aligns with the industry's shift toward bridging factory quality controls with real-time site operations.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop an autonomous monitoring agent using Mastra AI to predict battery degradation while tracking the carbon footprint of the computation with CodeCarbon.
How work is evaluated
The solution will be evaluated based on the accuracy of degradation predictions, the efficiency of the agentic workflow, and the integration of carbon tracking.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
Agentic Tool Validation
Ensures the Mastra AI agent successfully calls the 'DegradationModel' tool.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Prediction Mean Absolute Error
MAE for Remaining Useful Life (RUL) • target: 10 • range: 0-50
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Master Mastra AI's Workflow system to chain data ingestion, analysis, and notification tools.
Implement CodeCarbon decorators to monitor energy consumption during model inference on large BESS datasets.
Build a Battery Management System (BMS) anomaly detection tool using Scikit-Learn or PyTorch.
Design a RAG (Retrieval-Augmented Generation) system within Mastra AI to query technical battery manuals for troubleshooting steps.
Optimize agentic tool-calling to minimize latency and computational overhead in grid-edge environments.
Deploy a Mastra AI server that exposes endpoints for real-time telemetry ingestion and health reporting.
Reference links and supporting material
Sample data for 2 tasks
Challenge setup
The solution will be evaluated based on the accuracy of degradation predictions, the efficiency of the agentic workflow, and the integration of carbon tracking.
Challenge input
CSV file containing 50 cycles of voltage/current/temp data.
Agent execution
No agent tool is listed for this challenge.
Evaluated output
JSON containing predicted SoH (State of Health) and RUL in cycles.
- Ensures the Mastra AI agent successfully calls the 'DegradationModel' tool.
- MAE for Remaining Useful Life (RUL) • target: 10 • range: 0-50
- Prediction Mean Absolute Error target: 10
- 2 public reference cases
- Python execution harness
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
No tool records are attached.
Evaluation contract
- Agentic Tool Validation · Weight 1
- Prediction Mean Absolute Error · Weight 1
Metrics
- Prediction Mean Absolute Error · Target 10
Reference cases
- 2 public reference cases
Workflow tasks
- Task 1
Input: CSV file containing 50 cycles of voltage/current/temp data.
Output: JSON containing predicted SoH (State of Health) and RUL in cycles.
Execution harness: Python
- Task 2
Input: Execution log
Output: emissions.csv file with non-zero energy values
Execution harness: Python
Recipe freeze marker
No recipe freeze time is recorded.