Software factories hold the key to software-defined vehicles

Sep 5, 2023 by Karsten Hoffmeister, Dr. Ulrich Wurstbauer



In brief

  • The increase in complexity of SDV development is outpacing our improvements in efficiency 
  • MBSE and software factories can improve speed and efficiency while maintaining consistency  
  • To realize the potential of new technology and close the complexity-efficiency gap, we need a change in culture 


We need to change our development methods to tackle the complexity of software-defined vehicles.


Within the automotive industry, production is becoming more efficient. However, this linear growth in our production is easily outpaced by the rise in software complexity of software-defined vehicles (SDVs).  


As this gap continues to increase, it becomes ever clearer that if something doesn’t change, it will become impossible for us to progress to L3 and beyond. The complex systems of SDVs are a result of the increasing amount of software and the higher requirements for L3 — something that is only going to increase. So, as we cannot reduce the complexity, we must improve our efficiency. Our current methodology is outdated and not suitable for the complex systems we now have, so we need a new methodology. And this is how we’ll improve our efficiency.  


Model-based systems engineering


Model-based systems engineering (MBSE) uses models (rather than traditional documents and drawings) to show the different aspects of a system. As the models are represented using standardized languages, all parties get a holistic understanding of the system being developed. MBSE provides a structured representation of complex systems, thus facilitating analysis, simulation and validation. It also helps in the early identification of potential issues and makes it easier to manage changes to ensure system integrity throughout the system’s life cycle. 


MBSE and software factories can help


As mentioned, as we can’t reduce the complexity of SDVs, the challenge is to maintain consistency while achieving speed and efficiency. MBSE plays a vital role here by enabling communication and collaboration between suppliers and OEMs, MBSE helps achieve speed and efficiency. Software factories also increase speed and efficiency, but because they automate the software production process, they also maintain consistency. This combination of MBSE and software factories can be further boosted through modern development techniques such as AI (e.g., using AI to perform analytics). 


How do we use MBSE and software factories


So, combining MBSE and software factories provides the three essential elements of achieving speed and efficiency while maintaining consistency, but MBSE and software factories both come with their own set of requirements: 

  • Software factories need to use virtual validation technology to be virtual first' 
  • MBSE needs to move the contract API between supplier and OEM into the software layer  
    • The contract API needs to be moved up from the communication layer to the application layer 
    • To enable virtualization (of OEM middleware), we need a stable API between the application and the lower layers of the software this doesn’t need to be an industry standard across all OEMs, but needs to be stable for the specific vehicle/platform project 


Culture eats strategy for breakfast


Development culture needs a change. We have a lot of changes in technology, with the solutions to the challenge of software complexity available, so why is it still a challenge? Ultimately, it’s the way of working that needs to change.In the major automotive companies, this culture change doesnt get the attention it should be getting. We must have an intentional, planned cultural change to succeed in using the technology available to us — the technology that will enable us to close the gap between rapidly increasing software complexity and our efficiency. On a side note, purchasing culture also needs to change and auction-based purchasing is hindering real collaboration. 


End-to-end development


At Luxoft, we know the whole system of end-to-end development. We’ve got an unmatched, broad experience of clients and projects — giving us heterogenous systems knowledge. So, whichever systems you use, we can advise on and implement your development goals. Reach out to us to hear how you can close the gap between complexity and efficiency, while using technology that is already available. 


Watch Karsten Hoffmeister's presentation on how 'Software factories hold the key to software-defined vehicles' at the Tech.Ad 2023 conference. 


Karsten Hoffmeister , Vice President Autonomous Driving

Karsten Hoffmeister author linkedin

Vice President Autonomous Driving

Karsten is Luxoft’s head of Autonomous Driving. Previously, as the chief technologist for the same sector, he helped key clients in their transformation to develop software-defined vehicles. He has over 20 years’ experience in the automotive and software industries, this includes more than 14 years in technical management positions at world-class companies. He has published many technical papers and was also a key speaker at Tech.Ad 2022 and 2023.

Dr. Ulrich Wurstbauer , Chief Technologist Autonomous Driving

Dr. Ulrich Wurstbauer

Chief Technologist Autonomous Driving

Dr. rer. nat. Ulrich Wurstbauer is chief technologist for autonomous driving at Luxoft. He’s responsible for strategic developments in the field of autonomous driving, function development and virtual validation. Before starting his current in-depth work on automotive technologies with a strong focus on simulation and digital twin technologies, he received his PhD in the area of solid-state physics where he also did his post-doc. He worked on the newly developed 2D material graphene with its unique and very special quantum physics behavior. Ulrich has authored more than a dozen research papers in addition to filing four national and international patent applications.