Grid Power & Water Demand Co-Simulation with CrewAI & Fixie
Grid capacity limits and community opposition present key hurdles for new data center proposals. Build a collaborative multi-agent simulation using CrewAI and Fixie interfaces to model real-time electrical grid strain and water usage against regional utility capacity. Achieve sub-minute simulation accuracy across peak load scenarios.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop a multi-agent utility impact simulator using CrewAI agents and Fixie voice/text interfaces to model data center power and water footprint.
How work is evaluated
Evaluates power and water load projections calculated by CrewAI agents against standard energy physics benchmarks.
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.
Capacity Overload Identification
Correctly flags overload state when peak MW exceeds grid capacity
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
MW Calculation Accuracy
Percentage variance from theoretical thermodynamic baseline • target: 1 • range: 0-5
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Formulate CrewAI agent roles for Electrical Grid Analysis and Hydrological Modeling
Integrate Fixie interface endpoints for dynamic natural language scenario queries
Simulate hourly resource draw for 100MW+ data center configurations
Produce utility feasibility reports with actionable peak-shaving recommendations
Reference links and supporting material
Hourly grid load patterns and regional aquifer extraction limits simulated for high-density compute regions.
Challenge setup
Evaluates power and water load projections calculated by CrewAI agents against standard energy physics benchmarks.
Challenge input
JSON with facility_capacity_mw, cooling_type, ambient_temp_c
CrewAI
Enables multi-role agent collaboration
Fixie
Provides structured interface for conversational agent deployment
Evaluated output
JSON with projected_peak_mw, daily_water_gallons, grid_status
- Correctly flags overload state when peak MW exceeds grid capacity
- Percentage variance from theoretical thermodynamic baseline • target: 1 • range: 0-5
- Benchmark: GridSimBench
- MW Calculation Accuracy target: 1
- 1 public reference case
- Python execution harness
- Python sandbox (unavailable on Versalist)
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
- Fixie · Required
- CrewAI · Optional
- Adept · Optional
Evaluation contract
- Capacity Overload Identification · Weight 1
- MW Calculation Accuracy · Weight 1
Metrics
- MW Calculation Accuracy · Target 1
Benchmarks
- GridSimBench
- energy
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON with facility_capacity_mw, cooling_type, ambient_temp_c
Output: JSON with projected_peak_mw, daily_water_gallons, grid_status
Execution harness: Python
Evaluation tools
- CrewAI · Multi-agent framework
- Fixie · Agent UI & API Layer
Recipe freeze marker
No recipe freeze time is recorded.