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

We investigate what grade of sensor data is required for training an imitation-learning-based AV planner on human expert demonstration. Machine-learned planners [1] are very hungry for training data, which is usually collected using vehicles equipped with the same sensors used for autonomous operation [1]. 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 (see Fig. 1). 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 reason about why lower-quality data still provide good planning results.

Authors

• Long Chen
• Lukas Platinsky
• Stefanie Speichert
• Blazej Osinski
• Oliver Scheel
• Yawei Ye
• Hugo Grimmett
• Luca Del Pero
• Peter Ondruska

Keywords

• Training
• Road transportation
• Costs
• Sensitivity analysis
• Conferences
• Urban areas
• Training data
• Path Planning
• Low-quality Data
• Expert Demonstrations
• Data Quality
• Training Time
• Random Noise
• Intersection Over Union
• Domain Shift
• Autonomous Vehicles
• Small Amount Of Data
• Impact Of Quality
• Prediction Horizon
• Planning Approach
• Input Representation
• Modular Approach
• Longer Range
• Position Of Agent
• Human Drivers
• Sensor Configuration
• Imitation Learning
• Position Noise
• Variety Of Sensors
• Position Tracking
• Object Detection
• Transfer Learning
• Perceptual System