The automotive industry is undergoing significant changes with the increasing prominence of software and the complexity associated with it. This has led to the adoption of firmware over the air (FOTA) and software over the air (SOTA) approaches to ensure continuous updates in vehicles. Subscription-based models and cloud computing are playing crucial roles in the advancement of automotive architectures, allowing for remote activation of features and providing additional computing power for enhanced functionalities. In an interview with Rajesh Rajgor, Ashwin Ramachandra, Head of Digital Services (Transportation Business) at Tata Elxsi talks about the rise of software defined vehicles (SDVs) and over-the-air (OTA) software updates
Automotive original equipment manufacturers (OEMs) are prioritising the implementation of over-the-air (OTA) capabilities in current and future vehicle models. The shift towards software defined vehicles (SDVs) further emphasises the significance of firmware over the air (FOTA) and software over the air (SOTA), which are core technologies in the design, development, manufacturing, and ongoing usage of SDVs in shaping the industry’s future. Excerpts from the interview:
What is the next milestone that will change the way we look at future mobility?
The transportation and mobility industry are going through rapid transformation. There are several inflection points that will lead to significant future milestones. A primary driver is the evolution to a SDV-based electrical | electronic (EE) architecture. This will provide a platform for innovation for the next decade or so.
How will SDVs help OEMs drive revenue through subscription models and more?
Primarily, the basis of SDV is a strong FOTA | SOTA component. FOTA | SOTA provides the ability to update existing features and install new features after the vehicle is sold. This lays the foundation of a subscription-based revenue model for updates and upgrades. While FOTA | SOTA, as part of SDV, provides a framework for subscription monetisation, the actual features impacting the end-user need to be compelling enough for the consumer to pay a monthly or yearly fee. The FOTA | SOTA model forms the basis of how mobile phone software and applications (regardless of iOS or Android) are purchased and updated. The ease of use seen in the mobile phone industry needs to permeate into the automobile world.
How can cloud computing be effectively utilised in vehicles through SDVs?
Cloud computing is one of the significant inflection points in the transportation industry. There are several aspects of cloud computing that impact the experience of the end-user. Firstly, the concept of a connected car is more prevalent. This leads to realisation of several use cases such as trip planning, fuel efficiency (in the case of ICE), range estimation (in the case of electric vehicles), fleet management (for commercial vehicles) being live and in use today. Secondly, the FOTA | SOTA concept described earlier requires significant cloud computing and infrastructure to track and deploy updates. Thirdly, implementation of the entire subscription logic is done and maintained on the cloud. Further, cloud computing provides the foundation for creating digital twins of various automotive components such as battery, motor, infotainment, etc. that is the basis for AI or ML-based algorithms.
How is SDV driving simplicity while evolving towards a new architecture? What are the anticipated trends in the next 5-10 years?
The last few years have seen an exponential increase in software in automobiles. This increase has been fuelled by an increase in ECUs (basic computing blocks) and a further complication that arises due to the interconnection of these new ECUs. Both traditional domains like body, chassis and newer domains like ADAS, EV require significant computing resources to be realised. This has led to increased complexity and cost of creating platforms for automobiles.
SDV offers a paradigm shift by providing fewer but significantly larger capacity computing blocks (zonal compute, central compute).
The idea here is to reduce the high number of ECUs into a smaller set of zone plus central blocks, thereby simplifying the overall EE architecture of the platform. The challenge is now in identifying and isolation of workloads (for example, body and chassis computing, ADAS computing) and where they should reside (whether zonal or central or both). Once this challenge is addressed, the resultant architecture is significantly simplified and is scalable for future computing needs within the automobile.
What are the essential factors for SDV to become a reality?
SDV is already a reality today in niche segments. The adoption of this paradigm to the luxury, high-end and mid-range segments is expected to happen over the next 3-5 years. The key drivers are new and stronger silicon platforms, OEM-based middleware (there are several examples of OEMs developing their own versions of OS) and integration of cloud computing.
Can you elucidate the reasoning behind SDV for commercial vehicles, off-highway equipment, and fleet operators?
SDV, as a paradigm, will impact both commercial and passenger segments. SDV provides a computing platform where different computing workloads can run in parallel, providing user experiences. The composition of workloads in the passenger segment will be different from the workloads in the commercial segment. Hence, SDV is expected to significantly impact the commercial segment, just as it disrupts the passenger segment.
What is Tata Elxsi’s objectives regarding SDV implementation in commercial and passenger transportation in India and the overall revenue it will generate for the company?
Tata Elxsi is working extensively with OEMs globally and in India to consult and partner with the migration to SDV-based architectures. We expect SDV-based vectors to significantly increase over the next 3-5 years. However, SDV provides a platform for innovation in the transportation business for the next decade or so. Tata Elxsi continues to make investments to be positioned to benefit from the SDV revolution.