Improving Adaptiveness of AUTOSAR Embedded Applications

Hélène Martorell (1, 2) , Jean-Charles Fabre (3) , Matthieu Roy (1) , Régis Valentin (2)

1 LAAS - Laboratoire d'analyse et d'architecture des systèmes
2 Technocentre Renault [Guyancourt]
3 LAAS-TSF - Équipe Tolérance aux fautes et Sûreté de Fonctionnement informatique

AUTOSAR (AUTomotive Open System Architecture) is the most recent standard for automotive embedded systems. A major drawback of AUTOSAR lies in its lack of flexibility. Software-wise, ECUs (Electronic Control Unit) are tested, validated and uploaded; these three steps are performed in a monolithic process. Adding adaptability features into the standard is a major challenge in this context, and may result in consequent savings of time and money. On-line adaptation allows the inclusion of new functionalities in an efficient way. In this paper, we first extract relevant features from AUTOSAR for allowing dynamic software updates. Then, we define our approach for performing updates, and provide an evaluation of our approach on a RISC micro controller.

Mots clés: Partial Updates; Architectur; Automotive Embedded Systems; Real-Time Systems; AUTOSAR

Domaines: Calcul parallèle, distribué et partagé [cs.DC]; Systèmes embarqués

Hélène Martorell, Jean-Charles Fabre, Matthieu Roy, Régis Valentin. Improving Adaptiveness of AUTOSAR Embedded Applications. ACM Symposium on Applied Computing, Mar 2014, Gyeongju, South Korea. pp.384-390. ⟨hal-01062054⟩