Most of the current existing autonomous driving projects are essentially testing the increased complexity of the sensors and processing the vast amount of information from the cars, making the right decisions, and building the software algorithms. This processing requires massive software implementation. There is one trend, both in industry and academia, to consolidate clusters of more powerful application processors and accelerators into higher-performance multicore SoCs rather than discrete CPUs. Additionally, safety is very important in autonomous driving. This integration requires a dramatic change in the software architecture, resulting in a significant increase in the software footprint.
In this chapter, we mainly introduce how the safety features should be incorporated into the whole software design pipeline and introduce some tools from commercial companies to help the software design. Finally, at the end of this chapter, we would like to show one example of software architecture from academia.
Autonomous Driving Software Architecture
Autonomous driving algorithms and Its IC Design ; Kapitel : 12 ; 263-281
10.08.2023
19 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
High Reliability Autonomous Driving System Architecture
Europäisches Patentamt | 2020
|High Reliability Autonomous Driving System Architecture
Europäisches Patentamt | 2021
Software Architecture for Autonomous Spacecraft
NTRS | 1997
|METHOD AND APPARATUS FOR EVALUATING AUTONOMOUS DRIVING SOFTWARE BASED ON AUTONOMOUS DRIVING EVENT
Europäisches Patentamt | 2024
|REDUNDANT ARCHITECTURE FOR AUTOMATED OR AUTONOMOUS DRIVING
Europäisches Patentamt | 2025
|