RACECAR: Multi-Modal Sensor Data from Autonomous Indy Cars at 170 MPH

Multi-modal sensor data was collected from fully autonomous Indy race cars operating at speeds up to 170 mph (273 kph). Six teams contributed data from the 2021-22 Indy Autonomous Challenge, covering 11 racing scenarios across two tracks. The data is organized and released in both ROS2 and nuScenes formats.

Use Cases

• Developing and testing object detection and tracking algorithms based on the described LiDAR, Radar, and Camera sensor fusion.
• Exploring high-speed localization challenges based on data from banked tracks and high-acceleration scenarios.
• Training and validating autonomous driving models for overtaking and obstacle avoidance based on the multi-agent and solo lap data.
• Benchmarking mapping algorithms in dynamic, high-speed environments based on the described racing scenarios.

Strengths

• Data collected from real-world, full-scale autonomous racing at speeds up to 170 mph.
• Covers 11 distinct racing scenarios including overtaking, obstacle avoidance, and pit maneuvers.
• Data is provided in both ROS2 and nuScenes formats with a conversion library.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Six teams from the Indy Autonomous Challenge.

Collection Method

Data collected from fully autonomous Indy race cars during the 2021-22 competition.

Time Range

2021-2022

Freshness

unknown

Geography

Data collected across two unspecified race tracks.