How are ways of working evolving at Volvo Cars? And what can they learn from new players with a background in computer systems and software? To get some answers, we turned to Martin Hiller, Technical Leader Logical Elements and Software Architecture at Volvo Cars. Martin also shared his thoughts about when we will reach full automation (level 5) and he will join us at Vehicle Electronics & Connected Services with his presentation Going from a Computer-like Car to a Car-like Computer in Track A Electronics & Architecture during the second day of the event.
Martin, could you briefly explain your role as Technical Leader Logical Elements and Software Architecture at Volvo Cars?
Being a specialist at Volvo Cars means that you are perceived as somewhat knowledgeable within your area. As such you are expected to provide expertise and guidance within that area, as well as keeping a look-out as to what’s happening in the rest of the World and the rest of the industry within that area. The specialist track at Volvo Cars contains a number of levels, where “Technical Leader” is a step or two along the way. The higher up the ladder, the more senior and strategic the role.
My area, Logical Design Elements and Software Architecture, is a very broad area, and I’m actually not really sure I can easily explain the distinction between the various ingredients in the title. I use it as a way to work in the area of systems and software architecture of large distributed system, covering various views and viewpoints, among them the logical aspects and the software aspects.
In addition to hopefully bringing some expertise to the table I also work with ensuring that future directions and strategies in this area are, with each other as well as with directions and strategies in other related areas. Currently, this means heavy involvement in the definition of our next generation E/EA (electrical/electronic architecture).
In what direction are ways of working, processes and methods evolving in your field?
The dominating direction is without question going for an agile way of working, alongside necessary changes to organisation and responsibility split. This is combined with a wish to take more control of the product by increasing in-house development of strategic components, both software and hardware. Going forward, this may very well be the most profound change Volvo Cars has witnessed for a long time, maybe ever.
The automotive industry is moving towards more integrated computerisation, autonomy, and interconnection across vehicles, infrastructure and cloud. What is needed to succeed in this transformation?
After a few decades of steady evolutionary computerisation, we are now in a situation where it is not unusual for high-end cars to have over 100 different electronic modules connected across dozens of different networks, with most customer functions spanning many different electronic modules and networks.
The automotive industry is now at the brink of a profound transformation, a paradigm shift. The evolutionary process that has taken the vehicle from a purely mechanical structure to now being a computerized mechanical structure is now facing the final step: becoming a powerful computing platform with electro-mechanical peripherals situated in a mechanical housing – basically, a computer on wheels.
In addition to the changes internal to the vehicle, the product as such is transforming from consisting only of the physical vehicle to become a vehicle combined with surrounding services and features, provided by cloud enabled functionality, all provided in a seamless package.
This transformation will affect not only the product and the technology, but also organization, the way of working, the way different players in the industry relate to each other including commercial and contract related aspects. We will essentially have to become computer & software companies – a car manufacturer that cannot make this transition in time will not survive.
What can Volvo Cars learn from new players with a background in computer systems and software?
A term that is sometimes used is “the software defined car” – that is, the customer/driver experience is mainly defined by software. Seen from that perspective, it is clear that new players with a background in computer systems and software have a great advantage. They don’t see this as a transformation, they see it as a natural state of things.
If Volvo Cars, as well as all other incumbents, want to survive, we need to quickly adopt and internalise this philosophy that it is the software that defines the product, and realise that it is the software that differentiates us from our competitors, and that future customer expectations will be more akin to the experience they have with their smartphone – well-working software providing functionality that with regular over-the-air updates evolves over time.
Once we make this realisation, it follows that we need to improve our holistic aspect of the car’s electronic system instead of seeing it as a number of interconnected electronic subsystems. Then, a centralised computing platform with connected peripherals becomes a natural system structure. Following that, we need to allow for a substantial investment in this central computing platform that will carry the software that makes up our differentiation to our competitors, such that it allows for an evolving product.
In addition to the technological aspects, we need to acquire skills in efficient software development. This is underway with our agile transformation along with increased in-house development of software.
According to you, when will we reach full automation (level 5)?
The main difference between level 4 and level 5 is which conditions and driving modes the autonomous drive system can handle, where level 4 has limitations, and level 5 has not. The way I see it, I don’t think we will see a distinct step from level 4 to level 5, but rather see an extension of the conditions and drive modes level 4 systems can handle until they are, for all practical purposes, perceived as level 5 systems. Over time, the conditions and drive modes the system cannot handle will be increasingly uncommon and exotic. This would make the distinction between level 4 and level 5 more an academic exercise.
Divining when level 4 systems start approaching level 5 is difficult, and trying to predict this is like trying to predict the impact of modern smartphones when the iPhone was introduced. I’m sure we will see a tremendous development of autonomy systems once level 4 systems are introduced.
What are you most looking forward to by attending and presenting at VECS 2018?
By attending I hope to learn more about the ideas and directions of our friends and competitors, and by presenting I hope to get more feedback from innocent eyes on our ideas and directions. All tracks are interesting, so it will be hard to select which tracks and presentations to attend. Also looking forward to interesting coffee break discussions with like-minded (or not) people.