Duration: 24:11

98 views

1

Presented at SAAM Mobility 2021 (https://events.eclipse.org/2021/saam-mobility/)

Authors: Naci Dai, Deniz Memis and Burak Saglam (Eteration)

Abstract: Humans are very good at adapting to and driving in very different conditions and environments. Complex AI based software such as Autonomous vehicles (AV) are biased by the training data and they do not do well in different contexts. Environment perception and context awareness can be used to adapt and deploy complex networks of AI systems that are expected to perform safely on many different environments and edge cases. Typical Autonomous Drive (AD) software stacks provide localization, object detection and tracking using an array of sensor data. Additional sources of information such as V2X connectivity, passenger behavior, person-to-device mapping, urban context, degradation of traffic etc. complements the sensor data that can be used to describe a context. This context is used to predict and adapt the system of software and applications that are deployed to the vehicle.
Context Aware Software Stack for Mobility (CASSM) describes a model based approach for definition of a software stack for complex systems such as those found in autonomous vehicles. Machine Learning is used to detect a context and adapt the software stack, which can then be used to identify and deploy services to autonomous vehicles. CASSM will be an open source in-vehicle delivery of context-driven services and applications. This paper describes an architecture to define context-aware personalized & dynamically provisioned capabilities, applications & services to autonomous vehicles.