Modern autonomous systems increasingly span edge and cloud environments, demanding both real-time responsiveness and long-term analytic insights. Yet most cloud-native frameworks—designed for elasticity and scale—struggle to meet the strict latency, determinism, and data model consistency required at the edge. This talk introduces the cloud databus, a secure, extensible, and data-centric architecture built on the DDS (Data Distribution Service) standard that enables seamless, real-time, and scalable communication across the edge-cloud continuum.
This talk will introduce the architecture of a prototype cloud databus based on DDS. The presentation will show how this approach extends peer-to-peer DDS communication through the cloud without requiring protocol adapters or compromising real-time performance or Quality of Service (QoS) guarantees.
Key capabilities include low-latency observability and automatic bridging from constrained edge nodes to cloud-based systems. A case study on edge telemetry will be discussed, to illustrate how to simplify operational monitoring and debugging, offering advantages over general-purpose platforms like Datadog or Splunk.
Attendees will learn how the cloud databus differs from traditional edge-to-cloud integration technologies and why it is particularly suited for mission-critical autonomous applications. Benchmarks comparing this approach with MQTT, gRPC and Kafka-based architectures will be included.