Wipro – the vehicle digitization journey has no end point

Transformation from mechanical to software defined vehicles (SDVs) is redefining the entire automotive industry value chain, demanding new approaches to design and manufacturing.

Wipro estimates that cutting-edge AI and cloud technologies can reduce development time by 6-9 months and R&D costs by approximately 25%.

To find out more, Automotive Industries (AI) interviewed Dr. Swarup Mandal, Global Head Automotive, Wipro Engineering Edge and Thomas Mueller, VP and CTO, Wipro Engineering Edge.

AI: Where is the digitization of vehicles taking the industry?

Swarup: The changes brought about by digitization are a journey with no end point. Digitization is revolutionizing vehicle design, manufacturing, operation and servicing.

Design tools, simulators, and digital twins enable the design of the vehicle and its subsystems to be future-proofed. Doing it right first time at every phase substantially reduces the time and cost of vehicle development.

It also creates the potential for new business models and new revenue streams.

Thomas: Global OEMs have embarked on an evolutionary path to digitize their vehicles. Because the development of a new generation of electric/electronic architectures (EEA) can cost hundreds of millions of dollars, many such initiatives have been cut back or cancelled over the past 12 months due to significant changes in the global car market.

With the ending of EV subsidies and emission credit penalties in the US this year, an ever-bigger impact on the already slow sales of many BEV models is predicted. As a result, the massive multibillion dollar investments of Western OEMs into electrification over the last five to 10 years will defer the already late paybacks even further and shrink profits.

In contrast, many Chinese OEMs have taken a much faster, focused approach on digitalization and skipped some of the technical constraints, such as decades old technologies, in exchange for modern, cloud-based vehicle architecture that allows for faster, less costly development.

Development cycles for new vehicles and variants in China have reached impressive levels of technical, digital and on-road capabilities within a previously unthinkable timeline of less than two years.

Wipro helps OEMs compete by creating cars that get better every day. The Wipro Cloud Car architecture, pioneered in late 2020, was the first cloud-native EEA system to allow near infinite scaling of vehicle software, thus achieving a huge cost/performance benefit.

In particular, the area of driver assistance technology is a huge opportunity for most OEMs to create valuable, monetizable comfort functions such as “eyes on/ hands off” features.

The technical challenge is that only true end-to-end based AI systems can deliver real-world driving capabilities that match or outperform that of humans. To achieve this, an as yet unknown amount of AI compute power is needed in the vehicle.

For example, whilst just three years ago a then considered “huge” amount of AI inferencing performance of 1000 TOPS is today deemed insufficient to realize fully automated and autonomous vehicle driving.

AI models that aim to create autonomous driving functions require an enormous amount of real-world video and other sensor data to achieve the safety and performance to match humans.

Many OEMs cannot generate the required real-world data from their existing fleets.

Wipro’s Cloud Car “Harvester Stack” enables OEMs to “harvest” at fleet scale the data needed to train those world models for AI, but it will likely still take years to achieve the refinement of human driving and real-world interaction needed to gain full acceptancy. Tesla’s recent launch of Robotaxi in Austin/TX gives a glimpse of such capabilities coming from extremely powerful end-to-date vehicle AI platforms.

On this base, the hardware and software led innovations will have to continue at extreme pace to create true differentiation and real-world value. Standing still is definitely not an option for OEMs. The focus is to profitably sell existing and perhaps slightly refreshed products with powertrain options at price points consumers that appeal to consumers.

AI: How do traditional automotive companies shift from a hardware-focused culture to a software-centric world?

Swarup: Automotive OEMs have to change the approach for the design of electronics. Hardware needs to be designed to meet the requirements of software, as software has  become the differentiating factor both within OEM fleets and between OEMs.

Hence, OEMs are taking control of software more and more. Many OEMs have creating full blown software development centres or subsidiaries or forming exclusive partnerships focused on development, maintenance and servicing the software platform.

Their sourcing pattern of vehicle platform is changing. The traditional Tier-1 which used to supply hardware and software together, now provides just the hardware as per the specification given by OEM.

Software for the vehicle platform is being built in-house or by engineering service providers. In both cases, the OEM retains the intellectual property.

This drives reusability through harmonized design of software stack across vehicle lines and vehicle models.

Thomas: The best analogy for the OEM industry is perhaps to look at the handsets of the mobile phone market over the past 20 years.

There is no question that distinct design and differentiation matter for consumers. The success of design can be sign by many established brands such as Land Rover, BMW, Porsche etc. over “uniform” design found in many of the new Chinese BEV or NEV brands.

Consumer preference is influencing cockpit and interior design. Many consumers simply do not like the absence of haptic buttons on steering wheels or the absence of familiar column stalks.

Many OEMs have learned hard lesson about the difference between what they thought may be good technically, and what customers preferred. Sales numbers and customer feedback have led to many design re-thinks at major OEMs about how a digital user experience (UX) should be composed.

What matters to a customer is on most cases a “rounded” product design, which is the essence of hardware and software functions in order to create the best product experience possible.

Depending on the segment and positioning, vehicles embody certain emotions. Their powertrain, ambience, and looks appeal to buyers beyond standard metrics such as features and price points.

Recent polls show that, for many consumers, cars already have too many features. Price conscious buyers would hope to get a “no frills” car with as few features as possible to reduce the cost whilst meeting ever more stringent regulatory requirements.

The best recommendation in this regard is to create a product that makes sense for its user by blending seamless, boundaryless and frictionless digital experience with a smart, humanlike ability to support driving for those like to drive, and to automate driving for those who prefer to invest their time otherwise whilst in their car.

AI: What, in your experience, are the main challenges for companies and individuals?

Swarup: Automotive manufacturing tradition is being challenged to adopt the culture of the hi-tech industry. A software first approach is a paradigm shift from the design perspective. This requires transformation of the talent pool in terms of skillset and numbers of specialists.

It gets more complicated as the entire landscape of tools and technology stack for the vehicle platform are changing fast. This brings in additional challenges in organizational design and culture.

While manufacturing continues to be the key activity, keeping the core principles of automobile manufacturing and amalgamating it with that of hi-tech industry is becoming a key success factor for OEMs today.

The challenge lying ahead is the cross-pollination of mechanical and technological skills.

Thomas: The key challenge for all OEMs is to set aside traditional ways of engineering vehicle electronic architecture and to work from what the customer wants.

Car users expect a product that gets better over time and ideally learns from its use. Too many products labelled “smart” are actually quite the opposite.

Passenger cars are now embodied AI devices, at least to a small degree. And this will evolve rapidly, which means the OEMs’ ability to create full, end-to-end AI embodied products becomes imperative.

To give some examples: Bolting ChatGPT to a “Void Activate” button did not create much value for drivers and passengers, as this is not the way they prefer to interact. Instead, an LLM-based assistant like ChatGPT or Tesla’s Grok that ideally understands all vehicle functions including context, intention and reasoning, would be able to create value for both drivers and passengers.

We see in the market that imply “infusing AI” often falls way short of creating a true embodied AI product.

Many of our clients have started to take a completely different approach, including deep collaboration with key ecosystem new age partners, creating new leaner team structures, allowing “out of the box” developments such as the Ford “Skunkworks” or similar innovation initiatives.

Balancing the right approach and putting ever evolving AI in the car is perhaps the key to success. This is not about ever more screens that create little value and cause often more digital “friction,” but instead about creating a disruptive product that truly has the potential to be named the “iPhone on wheels.” In our view such a true innovation is yet to be created, despite many claims.

AI: Please tell us about the advantages of the Wipro Cloud Car ecosystem.

Thomas: Simply put, Wipro’s Cloud Car and its ecosystem are an EEA foundation open to our engineering clients globally to co-create a car that has the potential to become an “iPhone on wheels.”

Whilst this sounds very bold, we have created the hardware and software solution building blocks such as vehicle hyperconverged infrastructure to scale AI workloads seamlessly in cars, paired with our Shadow Mode and Harvester Stack to create true end-to-end ADAS and autonomous vehicle software that learns from its use and assimilates the human driver baseline.

In addition, our Cloud Car AI agents automate the ALM & PLM toolchains in a very synergistic way to reduce development cost and cycle times, whilst our Cloud Car virtual car allows us to create a phygital representation of every single endpoint in a vehicle to start simulating and virtually testing the entire vehicle hardware and software platforms.

This is not exclusive to passenger cars, but equally benefits the on-highway, off-highway and adjacent markets such as marine, aerospace and defense technology companies.

Swarup: It is now well appreciated that SDV development is a multi-dimensional complex task with a rapidly evolving technology stack. The heterogeneity in the progress of technology blocks needs a disciplined approach for adoption to realize the benefit of the advancement in terms of creating values of customers.

Wipro has built Cloud Car Solution which provides “chip to cloud” infrastructure with building blocks ready to be adopted for a quick industrialization of a vehicle platform. This infrastructure is built with ecosystem partners who are continuously investing in the domain. The solution is built with the following value propositions:

User Experience: This aimed to achieve the cabin experience hyper personalized with the use of AI/Agentic AI technology. The features also upgraded over the air in a user transparent manner.

Enhanced safety and security: As the number of connected features increases, safety and security are key aspects to be addressed throughout the life cycle of the vehicle. It helps to define the architecture and design principle for the vehicle platform.

Efficiency: Enable automakers to save costs, speed up innovation, and manage the entire vehicle lifecycle from engineering to aftersales on one platform.

Flexibility: The design and building blocks are focused on enhancing reusability which in turn helps to easily scale up and scale down the features easily based on the requirement.

AI: Where to for the technology and Wipro?

Swarup: Technology and automotive engineering service providers are moving fast toward digital, cloud-based, and software-driven solutions. The market expected to nearly double by 2030.

While technologies like AI, IoT, cloud-native, and connectivity are still maturing in the automotive domain, real-time analytics, simulation-based design, development and operation are a reality.

This helps automakers innovate faster, improve safety, meet tough regulations and enhance customer experience. Wipro’s partner ecosystem and solutions covering entire landscape “Chip to Cloud” and “Car to Plant” promised to facilitate the softwarization journey for the customers wherever they are.

Thomas: Wipro Engineering as a key global partner for product engineering-led innovation will continue to invest into truly differentiating solutions that help our clients to get the best out of their products and help them succeed.

Whilst AI is here to stay, we have a universe of opportunities to realize our vision of smart products jointly with our customers. We will continue to innovate as we have done throughout the 75-year history of the company.