Data drives the development of autonomous vehicles. During autonomous operation, an R&D vehicle gathers data from the sensors and produces logs. The raw sensor data, as well as onboard logs, is then used by engineers to improve onboard autonomy systems, including machine learning.
Onboard data is typically transferred to a large-scale storage device by fiber optics or by removable or swappable drives. Once the data is processed by engineers, it’s retained and used later for regression testing and engineering.
Data intelligence is an important part of autonomous vehicle research and development. Accelerate your R&D programs with high-performing, economical data-storage solutions, sensor truth benchmarking, simulation, data tagging and annotation, data-mining tools, mapping and more. Before you choose a solution, consider your requirements.
- Consistency or variation of sensor arrays
- Cameras/LiDARs per vehicle
- Your simulation and test strategy
- Duration of onboard data storage and transportation operations.
- Length of time in high-performance storage before archiving
- Number of engineers requiring access to uploaded data and their connection requirements
- Speed requirements for archived data retrieval
- Collaboration requirements with other sites
- Disaster reporting strategy
With 10 years of in-house experience in autonomy and advanced driver assistance systems (ADAS), combined with support from our expert partners, we can help you get up and running quickly. AutonomouStuff data intelligence solutions address common scenarios for ingesting, storing and archiving vast amounts of data from onboard sensors and systems.
Data generated by autonomous vehicle sensors (GB/day)
Autonomous R&D vehicles are fitted with a variety of sensors, which are used by onboard autonomy systems and machine learning algorithms. The raw data from these sensors is used to improve and train the onboard machine learning algorithms. Cameras and LiDAR generate most of the data.