The AutoSPADA Platform: User-Friendly Edge Computing for Distributed Learning and Data Analytics in Connected Vehicles

TL;DR

AutoSPADA addresses the challenge of turning edge-generated data from connected vehicles into timely, actionable insights by enabling analysts to deploy Python payloads directly to edge clients. The authors present a Go-based, distributed architecture that uses gRPC for user-server communication, RabbitMQ with MQTT for state updates, and Docker containers to isolate tasks, all under TLS and OIDC-based security. The platform emphasizes scalability, reliability, privacy, and resource efficiency, and provides Python libraries for both payload authors and end users. Evaluation on a Raspberry Pi client demonstrates interactive latencies suitable for rapid prototyping, underscoring AutoSPADA’s potential to accelerate edge analytics in automotive settings.

Abstract

Contemporary connected vehicles host numerous applications, such as diagnostics and navigation, and new software is continuously being developed. However, the development process typically requires offline batch processing of large data volumes. In an edge computing approach, data analysts and developers can instead process sensor data directly on computational resources inside vehicles. This enables rapid prototyping to shorten development cycles and reduce the time to create new business values or insights. This paper presents the design, implementation, and operation of the AutoSPADA edge computing platform for distributed data analytics. The platform's design follows scalability, reliability, resource efficiency, privacy, and security principles promoted through mature and industrially proven technologies. In AutoSPADA, computational tasks are general Python scripts, and we provide a library to, for example, read signals from the vehicle and publish results to the cloud. Hence, users only need Python knowledge to use the platform. Moreover, the platform is designed to be extended to support additional programming languages.

Figures (4)

• Figure 1: The logical AutoSPADA actors are the users (left) and the clients (right). The server nodes implement the infrastructure for transmitting tasks from users to clients and results from clients back to users.
• Figure 2: An overview of the AutoSPADA architecture showing the services, communication paths, languages, and protocols used in the platform. All communication is secured by TLS. Authentication between nodes is performed by OIDC using JSON Web Tokens (JWTs) for user nodes and through mutual TLS (mTLS) using X.509 certificates elsewhere.
• Figure 3: An entity-relationship diagram showing a simplified view of selected documents in the database, highlighting their relations. An assignment has many client-specific tasks. A task, in turn, has a payload and, optionally, a parameters document. NN is short for not null and E is for enum.
• Figure 4: Detailed view of the software components in an edge device running the AutoSPADA client. The major components are the sync loop, result handler (gRPC), signal handler(gRPC), Docker, and the WICE Signal Broker. A container supervisor thread is also started for each task but is not visualized with a component box.