Nakul Duggal: ‘Vehicle architectures are getting simplified significantly’
California-based Nakul Duggal, SVP and GM, Automotive, Qualcomm Technologies, in a freewheeling interview dwells on a host of issues relating to connectivity, electric vehicles, the chip shortage and the company’s outlook towards India.
What is your take on the massive transformation now underway in the automotive industry?
These are fascinating times for the transportation industry which is getting disrupted at various levels. There are multiple factors at play and one of the major ones is electrification, which nobody imagined would move at such an accelerated pace.
Electrification has brought an opportunity for automakers to reinvent themselves in terms of creating a new identity and how they associate with their customers. While it has also allowed newcomers to enter into the market, a lot of the existing car companies are even reinventing internally.
This reinvention is really about creating an experience with the consumer and also one with the car itself that now has to last much longer than what has traditionally been the case. Connectivity is the centre of it because if the car is not (connected) just like any other mobile device today, it really does not have any life once it’s driven off the showroom.
Moreover, as far as experiences go, we focus on so many different aspects of modernising the car and working with our customers to do that. It is the in-car experience, software ecosystem, and how the car is relevant for the use case it is designed for. So, a lot of catalysts have caused a massive amount of transformation in a very short period of time.
Why is connectivity the key driver right now and what role would 5G technology play?
One cannot really build a product today if it is not connected…and if it is not, you might as well not build it. And while we think of it as a very normal thing with respect to consumer electronics like a smartphone or TV, I do not think the vehicle is any different today, and that is really how software is being defined.
Furthermore, consumer expectations from their vehicles have changed drastically. So, if you think about safety, and being able to get access to a variety of services, one has to depend upon that layer of connectivity. And 5G is the latest wireless technology that we have been developing over the last several years. It is going to be here for the coming decade and more.
And that harmonised footprint will provide tremendous amounts of capabilities for new business models, and for really being able to sell the product very differently. I think the fundamental change that is happening is that the product that used to be a car or a two-wheeler was never really sold as a connected or a digital product. But if you start to sell it like that, how could you sell it without it being connected?
How crucial is cybersecurity and with increased connectivity, what is the role of data going forward?
Security is paramount for all connected devices and to that extent, a variety of bodies have taken active steps to make sure that there are standards around cybersecurity. The UN 21434, for instance, is the standard that focuses on ensuring that consumer security requirements are available broadly across the overall vehicle platform.
It also makes sure that it gets into the recommendations around how subsystems should be architected. So, there are standards in place for technology companies to be in compliance with and the system architectures are also shifting to adopt these safety standards.
As regards data, the car, which today is a sensor on wheels, has access to a tremendous amount of information at a global footprint. This gives birth to a massive responsibility at the automakers’ end to ensure privacy of the consumer data.
One has to create an approach and segregate data, based on what can be used to improve the product, or what could be shared with suppliers for them to build the product better. One has to draw the line. While not much is talked about, there’s a lot going on in this space — for instance, Qualcomm is getting actively involved in advanced driver assistance systems (ADAS), which is all about data.
In that sense, clear partnerships are emerging between customers and technology providers like Qualcomm in order to get access to data. I think we will see a lot of maturity in the business ecosystem when it comes to data sharing in the coming years.
What are the innovations that Qualcomm is working on 5G technology?
Our product portfolio has actually evolved over the last 10 years and we now build what we call the Snapdragon Digital Chassis, which is connectivity for the car. While e 5G is one of those pieces, we also built Wi-Fi, GPS solutions, power-line communications, as well as Bluetooth for the vehicle.
We started to focus on the in-car infotainment experience about eight years ago and are now market leaders in terms of all premium infotainment systems for the entire global automaker ecosystem. Recently, we have doubled down on the ADAS space.
We were earlier doing semiconductors for ADAS technology and in April 2022, we completed the acquisition of Arriver from Veoneer which brings us the entire stack. Therefore, we have a very broad portfolio of capabilities.
How are your relationships with OEMs on technology development for electrified and autonomous driving?
The business model with automakers has evolved quite a bit as well. One thing that OEMs have realised in the last 5-7 years — as the competition really heightened for them and as electrification became important — is that they started to compete with a new set of players which didn't really have the legacy or the burden of having to run a combustion engine business.
So, a lot of the new players started off with a clean sheet design, and a software-first approach. What became very apparent to automakers was that they needed to have technology partners that they could rely on and build long-term relationships with, such that they would not have to think as much about how to modernise the platforms that they are building as opposed to the traditional Tier 1-based model, which calls for a new design every three years.
Today, most of our prized relationships with automakers are directly with them. Of course, we work with Tier 1s because even they have a very important role to play, but we work in a triangular relationship where the OEM is at the top and Qualcomm, as well as the Tier 1 work together, but directly with the OEM to be able to make these things happen.
I think it has worked quite well, because what we are seeing is that automakers are bringing in-house, a lot of capabilities that they were relying on Tier 1s for. This allows them to invest in a platform that lives much more than one generation.
It makes a huge difference because we too are able to partner with the automakers on a multi-generational basis and think about continuity in our designs, as we now focus on selling our entire digital chassis, as opposed to looking at the opportunity component by component.
Will this proximity of technology providers to OEMs pose a risk to the existing Tier 1 ecosystem?
If the car is modernising and at a pace that it has to keep up with the smartphone, one needs to have access to partners that can move at that pace. And the way that I would describe it is that the role of the Tier 1s is to really be a very complex system integrator, which they still are. While in the past, automakers used to rely on Tier 1 suppliers for bringing in differentiation, today they are experimenting with potential business models around differentiation and finding more residual value once the vehicle is sold.
This mandates the automaker to be directly involved in the conversation, as opposed to leaving it to a very traditional ecosystem. And I think Tier 1 suppliers are still going to be very relevant in areas where it doesn't make sense for the automaker to change anything.
However, in cases where automakers are looking to own the intellectual property, they would work with the technology partner directly, which is really true for any ecosystem. I think it's just a sign of the times.
Where are vehicle architectures headed in terms of process node technology of the semiconductor chipsets?
As the car architecture started to shift to where automakers wanted to change the experience and do more complicated things like autonomous driving, connectivity, higher-end infotainment and integration of subsystems, companies like Qualcomm entered and offered to bring a lot of technology. So, in some sense, technology companies actually ended up accelerating the pace at which the auto industry could move.
And for us, one of the things that we have to do is to keep advancing on the process node, because that is the only way that one can actually deliver a complex product like a smartphone. So, the two things that are a big advantage for us is, one, access to a massive amount of technology, and the other is to be forced to move the semiconductor pace to the next process node.
Since the car is looking to modernise very quickly, we took the approach of using our latest-generation smartphone chipset products in automotive. The vehicle of the future cannot afford to be relying on 10-15 different process nodes right from 40 nanometre to 7 nanometre, and we can actually bring all of this functionality into two or three systems on chips (SOCs), and use software to really bring the differentiation.
Therefore, vehicle architectures are getting simplified significantly from a very large number of smaller controllers to a small number of more powerful controllers. And this shift modernises the vehicle and gives an opportunity for the automaker to reset the architecture. This means one would not have to pick older process nodes and can actually move to a new process node and makes the architecture very software-centric.
What direction do you see the industry going after the prolonged chip shortage?
Car architectures are not that uniform as we have a lot of different automakers building vehicles using various vintages of chips. And the crisis has affected different suppliers, foundries and processes in different ways. It is a complicated situation as there’s a geopolitical aspect to it, availability of skill set, tools, and investments.
Over the next two to three years, we will start to see a transition towards a more modern vehicle architecture which will be on a modern process node, backed by a modern supply chain. But there will also be a window of time, during which the legacy architectures will have to be supported as well. These decisions were made many years ago, and therefore, currently there is a dependency on a variety of process nodes.
The big challenge that obviously exists, if one is on an older process node which does not have a lot of demand, is that it costs a lot more to keep that architecture around. I think what automakers are struggling with is how much of their car architecture is going to be able to move to a newer generation, and how much will likely stay on an older one, and more importantly, how would they go deal with that gap.
We are fortunate because we have a very large semiconductor business. So, we were able to manage our automotive demands quite well through the chip crisis, and we did not really have much of a challenge in meeting those demands. But I think, every automaker that we work with has become wiser up to what their complexity is and they're all moving very quickly away to not find themselves in situation again,
Having said that, I do not think it is the same problem for every automaker. The ones that have a very large portfolio of vehicles, and are very globally deployed, as well as sell into many different tiers of vehicles, while being low on flexibility; I think they will probably have to be much more careful.
When do you expect the chip shortage to end?
While it is hard to predict something like this, I can say that from our perspective, we are certainly out of the woods. We were out of it almost a year ago because we were able to leverage our scale, and given our priority to the automotive business, we ensured that we were not the long pole for any of our customers. But I think the problem statement is a lot more complex depending upon which part of the value chain one is talking about.
Do you see scope for making these semiconductor chipsets in India?
There is huge complexity behind it as one needs scale, talent, environment, skill set as well as the business case. While there is a lot of focus on subsidies being available from governments around local manufacturing, I would imagine it to require scale not just from a local set of customers but also maybe from a regional set of customers. I am sure those are all opportunities that would be available for semiconductor suppliers that do their own manufacturing.
How is India contributing to Qualcomm’s global scheme of things in product development and innovation?
There is a significant need to find the ability to differentiate the next-generation platforms, which are being designed for multiple applications and use cases – both for automotive and consumer electronics. And to me, it really is about how the automaker starts to think about building a platform that is modern yet flexible, and one that is able to accomplish a lot of differentiation while being on a new architecture.
I think that there is a tremendous amount of capability in India. Qualcomm has a very large base in the country where we do an immense amount of hardware and software development not just for India but also for global markets. There is really no dearth from a talent pool perspective.
'We are willing to disrupt the established insurance norms': Shanai Ghosh, CEO & MD, Zuno General Insurance
For digital-first Zuno General Insurance, 2023 was a year of significant milestones and impressive growth, marked by a s...
FADA President Manish Raj Singhania spoke with Autocar Professional about the auto industry outlook for the coming year ...
Pratap Bose, Chief Design and Creative Officer of M&M Auto and Farm Sector discusses the role of creativity, technology,...