Ashwin Ramachandra has been associated with Tata Elxsi since April 2021 as the head of its digital wing. In his career spanning over 29 years, he has worked with Texas Instruments as a software engineer and with Sasken Technologies as Vice President of Technology. Interestingly, he started his professional innings with Tata Elxsi as a Software Engineer in January 1994.
Regarding Software Defined Vehicles (SDVs), what role does Tata Elxsi play?
We are involved in programmes with OEMs and Tier-1s, but certainly, we are in customer programmes in various shapes and forms, contributing to how SDVs will be part of OEMs' roadmaps in the coming years.
How are the OEMs reacting to the concept of SDVs? Do you think this is just a buzzword?
One perspective is that the new-generation OEMs have an SDV-first approach. These are the folks who have no legacy and are more into EVs. However, existing OEMs, who have a legacy of, for example, internal combustion engines, etc., have a roadmap of transitioning from current architectures to SDV architecture.
So either way, whether it is new age OEMs that we talk about or legacy OEMs that are migrating into this newer paradigm, we see the firm conviction on the roadmaps that they put out to us and the rest of the world. The conviction is definitely among the top OEMs to migrate to one SDV architecture. It is something that will happen over a period of the next three to five years.
Why are the new-generation OEMs finding it easier to transition?
It is about the application of resources. For example, if I had an R&D resource today and had no existing commitments to put out vehicles this year or the next year, I would dedicate that R&D resource entirely towards the cause of realising an SDV.
A lot of these new-age OEMs have no existing commitments. They don't have cars on the road. They don't have service obligations on the road, etc., so SDVs are simpler for them.
Do you see SDVs as more oriented towards luxury and premium segments?
Five years later, the car that one will buy will probably be the least-best version of the car. From that day onwards, for the OEM, it might be possible to provide features, either as a regular update free of cost or on a subscription basis, month on month.
What SDV does is it enables a platform or a mechanism for OEMs to get continuous revenues, not a one-time sale. If you apply that across the car segments, it is even applicable to a $18,342 to $24,456 (INR15 to INR16 Lakh) car as well. Apart from the entry-level segment, SDVs will be deployed in every other segment.
What is the trickle-down time for SDVs from premium to mass-segment vehicles?
The platform approach drives this, and in a lot of cases, sedans and hatchbacks are based on the same platform. OEMs retail feature-rich vehicles and feature-less vehicles based on the same platform. However, the architecture does not change.
SDV makes it even more amenable to this sort of mechanism, as the same is a hypercomputer platform with the potential to run a lot of computing. We believe this kind of platform architecture enables differentiation to a great degree. You could buy a bare-bone model with almost nothing in it, and then I could subscribe to a few features. The trickle-down time will depend a lot on OEMs and the end consumers.
Will SDVs also increase revenue opportunities for OEMs?
Besides reducing complexity, which is an extremely important driver for OEMs towards SDVs, there are five or six critical drivers. Today in a luxury car, there are at least 50 to 100 plus controllers, but when you move to an SDV, that number will drop drastically. Via SDVs, managing the electrical architecture and the software on the car becomes a little bit simpler. Hence R&D investment in creating multiple variants and products would go down. Premise number two is OEMs looking at consumers as extended revenue models, not just a buy-and-forget type of a thing,
Do you see a grey area that might hinder the growth of SDVs?
The downsides are not completely clear at this point. There were certain architectures that traditional OEMs were very comfortable with, but today's architectures are changing quite differently. For example, a small ECU on a canvas will give way to a zonal controller on an automotive Ethernet.
There is certainly the element of the unknown, and it will need the risk-taking ability to step out of the comfort zone; however, SDVs are inevitable because of the advantages.
When would the majority of OEMs showcase something substantial around SDVs?
By 2025 we would see leads and concepts, some of which might also go into production. This would be barring the folks catering to entry-level. However, a 2025 or 2026 architecture could be reused in a 2030 vehicle for the entry segment.
That way, the costs of R&D are already amortised. In another eight years, it's quite likely that even entry-level cars will get connected.
Do you see the commercial vehicle industry adopting this first or the PV industry?
There are efforts on both sides because OEMs in both segments are driven by the same considerations: cost, complexity reduction, and revenue models beyond a one-time purchase. Both industries are certainly investing in this as an initiative, but PVs are more likely to come out ahead on this.
Also Read