What data do we need for training an AV motion planner?

TL;DR

This paper demonstrates that training autonomous vehicle motion planners with lower-quality sensor data can outperform high-quality data, enabling cheaper data collection and broader access to imitation learning methods.

Contribution

It shows that lower-quality sensor data can be effectively used for training AV motion planners, reducing costs and increasing data availability.

Findings

Lower-quality data outperforms high-quality data in planner training.

Cheaper sensors can be used without sacrificing performance.

Sensitivity analysis reveals key factors influencing planner performance.

Abstract

We investigate what grade of sensor data is required for training an imitation-learning-based AV planner on human expert demonstration. Machine-learned planners are very hungry for training data, which is usually collected using vehicles equipped with the same sensors used for autonomous operation. This is costly and non-scalable. If cheaper sensors could be used for collection instead, data availability would go up, which is crucial in a field where data volume requirements are large and availability is small. We present experiments using up to 1000 hours worth of expert demonstration and find that training with 10x lower-quality data outperforms 1x AV-grade data in terms of planner performance. The important implication of this is that cheaper sensors can indeed be used. This serves to improve data access and democratize the field of imitation-based motion planning. Alongside this, we perform a sensitivity analysis of planner performance as a function of perception range, field-of-view, accuracy, and data volume, and the reason why lower-quality data still provide good planning results.