Software Defined Vehicle - this is not a hype.

Last week I got asked whether the current Software Defined Vehicle will not be a hype like the autonomous car or the connected vehicle?

Let me go back to the year 2005 - when I presented my thoughts in a keynote to the 2005 Embedded World Conference. Actually, the ideas were from summer 2004, when between jobs I had the opportunity to elaborate and extrapolate the trends I was seeing towards the year .. 2020!

In 2005 I foresaw the rise of software content in motor vehicles. Slowly but surely all mechanical, electrical, mechatronic and electronic functionality in vehicles would be superseded by software and software would become the major cost factor in vehicle electronics:

When in 2005 only 5% of vehicle recalls were due to software, for the year 2020 I foresaw around 25% and pointed to strategies to combat those tendencies. One of these was AUTOSAR – founded in 2003 as a successor to previous modular software initiatives – in 2005 AUTOSAR gained international popularity. Also the use of hardened RTOS in vehicles started in the infotainment and telematics domain, where safe and secure operation was mandated by connectedness to “open” networks in the outside world.

Thanks to company sponsor I contributed in AUTOSAR Work Packages Methodology and MCAL – the later was my specialty since I was microprocessor expert for embedded systems (MC 68K, PowerPC, ARM). We established AUTOSAR classic, then, and around ten years later, mid-2010s the need for more dynamic configuration and less static interfaces between the then reigning plethora of ECUs became obvious – AUTOSAR Adaptive was born and I participated in the adaptation of QNX RTOS towards that goal.

Now in the 2020s the world of big IT – cloud and network provider, continues to integrate mobility. Private motor vehicles are low hanging fruits as the sale of cars is dominated by status, comfort, and convenience, when utility is more of an afterthought. And individuals are more likely to pay a princely wage (“fürstliche Entlohnung”) to the suppliers of advantages that software defined vehicles could promise.

On a first glance some of these promises seem ridiculous, like subscriptions for heated seats, but others .. say, if for a one-time payment bugs get fixed and security holes get filled – many of us would pay a monthly fiver or even ten to keep their vehicle up-to-date. Multiply this by a million vehicles this enables the continuous improvement even for producers of low margin mass market vehicles. When our smartphones don’t get any updates after two or three years, our cars will!

See also my forthcoming “BK Travel Guide to the Software Defined Vehicle”

And my previous article “Software Defined Jungle”: https://www.dhirubhai.net/pulse/software-defined-jungle-bernhard-kockoth