AI-Driven Optimal Optical Ground Station Placement for LEO Constellations
Design and implement an AI-powered system to determine the optimal placement and operational strategy for a network of optical ground stations (OGS) serving a large Low Earth Orbit (LEO) satellite constellation. The challenge involves maximizing data throughput, minimizing latency, and ensuring resilience against adverse atmospheric conditions (e.g., cloud cover, turbulence). Participants will leverage modern machine learning and optimization techniques to model satellite-to-ground links, simulate atmospheric interference, and intelligently schedule data downlinks. The solution should demonstrate how to dynamically adapt OGS operations based on real-time environmental data and satellite availability. This challenge emphasizes practical application in space communication, where maximizing data return from burgeoning LEO constellations is critical. It requires integrating orbital mechanics, atmospheric modeling, and advanced optimization. Success will be measured by the network's ability to achieve high data volumes under realistic, variable conditions. The use of robust MLOps practices is encouraged to manage the iterative design and evaluation process.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop an AI-driven optimization framework using Command R+, Metaflow, and Ax for resilient optical ground station network design.
How work is evaluated
The evaluation will focus on the performance of the optimized OGS network (data throughput, resilience, latency), the robustness and scalability of the Metaflow workflows, and the effective integration of Ax and Command R+ for decision support and optimization.
Challenge datasets
Public datasets, plus the private datasets you have access to.
What you should walk away with
Learning objectives will be added soon
Use the overview and evaluation guide as the source of truth for expected outcomes.
Reference links and supporting material
Sample data for 2 tasks
Challenge setup
The evaluation will focus on the performance of the optimized OGS network (data throughput, resilience, latency), the robustness and scalability of the Metaflow workflows, and the effective integration of Ax and Comma...
Challenge input
{'num_ogs': int, 'ogs_locations': [{'lat': float, 'lon': float, 'alt': float}], 'constellation_definition': dict}
Agent execution
No agent tool is listed for this challenge.
Evaluated output
{'total_data_gb': float, 'avg_latency_ms': float, 'link_availability_pct': float}
- Ensure the Metaflow pipeline executes without errors and generates a final optimized configuration.
- Verify that Ax has been successfully integrated and run for a significant number of iterations to find an o...
- Total Data Throughput (GB) target: 3000
- 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
- Ensure the Metaflow pipeline executes without errors and generates a final optimized configuration.
- Verify that Ax has been successfully integrated and run for a significant number of iterations to find an optimal solution.
Metrics
- Total Data Throughput (GB) · Target 3000
- Average Data Latency (ms) · Target 100
- Link Availability (%) · Target 95
- Command R+ Utility Score · Target 7
Workflow tasks
- Task 1
Input: {'num_ogs': int, 'ogs_locations': [{'lat': float, 'lon': float, 'alt': float}], 'constellation_definition': dict}
Output: {'total_data_gb': float, 'avg_latency_ms': float, 'link_availability_pct': float}
Execution harness: Python
- Task 2
Input: N/A
Output: {'workflow_status': 'success'|'failure', 'optimization_iterations': int, 'best_config_found': dict}
Execution harness: Docker
Recipe freeze marker
No recipe freeze time is recorded.