Driving Innovation In Autonomous Vehicles – Indy Autonomous Challenge Blog

The data-centric connectivity architecture is based on a software databus that is designed for exceptionally complex systems like autonomous vehicles. The databus operates as a shared application space where all of the data lives. This simplistic design is also what allows applications and devices to work together as one integrated system. The databus distributes high-volume, high-speed data in motion across APIs — unlike a database, which manages data at rest. Data serves as the interface between devices but isn’t necessarily stored anywhere. Messages do not need to be sent through brokers to access or process that data. The databus architecture allows for real-time response (or near-zero latency), because it is able to move high volumes of data-in-motion from multiple sources to wherever it is needed, without retrieving messages from storage.

In AV environments, this provides the structure to rapidly move data between the components of the self-driving software to tune, record, replay, visualize, analyze and iterate the data flows to extract every millisecond of race-winning performance from the software.