Extending Automotive SPICE 3.0 for the use in ADAS and future self‐driving service architectures

Abstract

The SOQRATES (www.soqrates.de) working party has been established in 2003 with the support of the Bavarian SW initiative. Major automotive suppliers joined forces to exchange best practices in topics such as Automotive SPICE, functional safety, and cybersecurity.The research method of SOQRATES is to compare the best practices, and in case that a specific design pattern is accepted by all parties, it is declared and published as a state of the art.Some of the results of the working party have been packaged into training courses.For example, in the EU project SafEUr (518632-LLP-1-2011-1-AT-LEONARDO-LMP, 2011-2012) a European partnership with inputs from SOQRATES developed a skill set, training materials, and best practices for the implementation of ISO 26262.For example, in the EU project AQUA (Knowledge Alliance for Quality in Automotive, EAC-2012EAC- -0635, 2013EAC- -2014, a European partnership with inputs from SOQRATES developed a skill set, training materials, and best practices for integrating Automotive SPICE, ISO 26262, and Six Sigma.For example, in the EU project AQU (Automotive Quality Universities, 2015-1-CZ01-KA203-013986, 2015-2017), a European partnership with inputs from SOQRATES applied the AQUA concept with universities in Austria, Germany, France, and Czech Republic who educate students that will work in the automotive industry.Also, the working party elaborated integrated assessment models where the Automotive SPICE 3.0 has been merged with ISO 26262 (further safety related questions) and SAE J3061 (further cybersecurity questions). This paper will look into the future of self-driving cars and discuss the design patterns that are currently analysed in the working party to support a vehicle in future self-driving infrastructure architectures and processes. KEYWORDSAutomotive SPICE, cybersecurity, functional safety, ISO 26262, SAE J3061, service architectures for automotive 1 | INTRODUCTION ADAS stands for Autonomous Driving Assistance Systems and realises functions that support the driver but still keep the driver in the flow. It is still expected that there is a driver with a driver licence who is part of the control flow. From 2030 on the plans from OEMs are to produce self-driving cars where there are no drivers, the passenger is a person that is provided with a mobility service. The car itself must control the situation (supported by infrastructure), and also the insurance model will have been changed by then. Cars will have a black box that logs all vehicle data from all electronic control units (ECUs) and insurance will go with the car and the component, and will not be on the driver as a person any more.In case of the steering example in this paper, the authors also outline the future of a self-driving scenario.Automotive companies experience an exponential increase in functional development. Major car manufacturers develop vehicle functions that can be decomposed into features (functions) on ECU and supplier level. A real time communication (via a bus) of a set of electronic control units ...

The impacts of the COVID‐19 pandemic and the globalization on how automotive assessments will be performed in the future

“…However, due to the distributed structure of automotive projects where expert teams working in one project come from different countries, it is expected that the 57% will further grow. The car functionality complexity will further grow in the future, 5,13–15 which will lead to more integration of experts from different places in the future. An online assessment strategy allows to connect these experts from different locations.…”

Process improvement guidance for successful automotive SPI implementation

“…Principle 'use dynamic and adaptable models as needed' Do integrate different norms into one V-model so that with one assessment and improvement all norms can be handled. [5][6][7][8][9][10]42,43 Do not create thousands of templates and one template per different norm, letting projects drown in formalism. Do define a structured portfolio of functions (see Figure 3) and map them to customer project releases.…”

“…Do integrate different norms into one V‐model so that with one assessment and improvement all norms can be handled 5–10,42,43 …”

Cross-Cutting Approach to Integrate Functional and Material Design in a System Architectural Design – Example of an Electric Powertrain

“…the weather conditions. This means that all these registered data are connected to each other which implies that the car becomes a data cloud [5,6,18,25,26].…”

“…Volkswagen defined vehicle functions (FUN principle for function based vehicle development) and each supplier maps their own subfunctions/features, and system requirements to these vehicle functions [11,23]. Automobile manufacturers apply assessments to check the traceability, safety, and security of those functions by assessments [12][13][14][17][18][19]. The fulfilment of the related norms requires the definition of such functional designs, effect chains and dynamic views.…”