Generative Simulation Benchmarking for Autonomous Urban Air Mobility Routing for Low‑Power Autonomous Deployments

Описание и применение

Генеративное моделирование сценариев позволяет создавать реалистичные, противоречивые условия для автономных летательных аппаратов, обеспечивая оценку маршрутизации с учётом ограничений мощности и времени.

Задача

Low‑power autonomous agents (drones, ground robots, embedded AI) fail when deployed because simulators are too clean and do not reflect real‑world constraints such as wind, sensor dropout, compute limits, and regulatory changes.

Результат

The generative benchmarking loop produces realistic, adversarial scenarios that expose routing policy weaknesses early, ensures policies are evaluated under deployment‑level precision (e.g., INT8), and quantifies energy consumption per decision, guiding design toward the optimal trade‑off.

Рабочий процесс

• 1. Train a conditional VAE to learn the manifold of realistic UAM scenarios (wind fields, traffic density, sensor noise, regulatory masks).
• 2. Sample scenarios conditioned on difficulty from the VAE.
• 3. Run the routing policy on each sampled scenario on hardware‑mirrored compute and energy models.
• 4. Measure success rate, latency, and joules per decision.
• 5. Aggregate results to assess Pareto frontier between capability and consumption.
• 6. Iterate: refine scenario generator and policy based on continuous benchmarking feedback.

Доказательства

The simulator I'd used was too clean, too generous, and utterly disconnected from the constraints of low‑power edge hardware.

Источник

Every benchmark run must report joules per decision, not just success rate.

Источник