Robotics/Autonomy/AUV/UAV Field Testing & Benchmarking

SOFIA SAT CLUB can contribute a focused sub-task for robotics/autonomy projects: structured testing, benchmarking, and (where feasible) dataset collection in real outdoor conditions. Our program includes “autonomous robotics challenges,” engineering problem-solving workshops, and capability for rapid prototyping (e.g., 3D design/printing), making us suitable for verification/testing work packages and field evaluation tasks.

We can support autonomy evaluation themes such as terrain traversability, low-cost sensor fusion, edge inference constraints, robustness to weather/lighting, and logistics-like scenarios (route planning, waypoint missions, inspection runs).

Scope / activities:

• Define test scenarios + success criteria (repeatable routes, obstacle sets, perception targets)

• Field testing in representative terrains (urban edges, parks, uneven ground, low-visibility conditions when possible)

• Benchmark autonomy modules (localization, perception, planning, safety stops, energy use)

• Create evaluation artifacts: logs, summary metrics, “what failed and why,” and recommended improvements

• Optional: produce a small dataset (images/video + sensor logs + annotations) for training/benchmarking

Expected inputs from consortium:

• Target robot platform or simulation environment (if they want sim-first)

• Baseline autonomy stack, logging format, required metrics

• Safety requirements and test constraints

Deliverables we can own:

• Test plan + risk/safety notes

• Benchmark report (metrics, failure modes, comparisons across iterations)

• Dataset package (if agreed) + documentation

• Recommendations for improving robustness / deployment readiness

Typical effort / size: 0.5–2 person-months
Keywords: autonomy, robotics testing, benchmarking, dataset, field validation, robustness

We are happy to join first as a small task partner and scale our contribution once the consortium structure is confirmed.