Time-Triggered Organic Computing Architecture for Autonomous Driving Vehicles Using List Scheduling

Konferenz: AmEC 2024 – Automotive meets Electronics & Control - 14. GMM Symposium
14.03.2024-15.03.2024 in Dortmund, Germany

Tagungsband: GMM-Fb. 108: AmEC 2024

Seiten: 6Sprache: EnglischTyp: PDF

Autoren:
Qosja, Mario; Meckel, Simon; Obermaisser, Roman (Chair for Embedded Systems, University of Siegen, Germany)

Inhalt:
As the autonomous vehicles market is expected to grow in the future, their functionalities will increase too, leading to complex embedded computer systems. To address this, organic computing has emerged as a research area that takes inspiration from biological entities to handle complex distributed embedded computer systems. Organic computing has improved adaptability and robustness, while also reducing development efforts. However, it has some drawbacks in terms of determinism, composability, and dependability, which are key features for safety-critical applications. Distributed computer systems using time-triggered communication networks possess these characteristics and thus show distinct advantages for safety-critical systems. By combining artificial DNA and hormone features with time-triggered communications, we can make these systems safer, more reliable, and suitable for safety-critical applications. Therefore, we present a time-triggered organic computing architecture where the time-triggered schedule is produced by a list scheduling algorithm during runtime. All the tasks on the systems are executed according to the computed schedule. To demonstrate the concepts and the system model, we performed evaluation test in a simulator in which the system executes tasks according to predefined schedules. The evaluation of use cases shows improved temporal predictability and fault containment.