Today’s vehicles feature advanced technology, yet the automotive industry is just starting its shift towards Software Defined Vehicles (SDV), aiming to innovate, personalize, and foster collaboration for profitability, writes Daniel Elhs, Business Director and Subject Matter Expert for SDV, Valtech Mobility GmbH and Nilesh Chauhan, VP, Automotive and Mobility, APAC and MENA Region, Valtech Mobility GmbH
The automotive journey so far
Automakers have spent many decades to achieve the current level of quality in their products by combining the experience and expertise of various engineering fields such as mechanical engineering, electrical engineering, material science, manufacturing engineering and so on.
And as a result, we see today’s vehicles loaded with lots of technology and features. But the current times are just the beginning of a greater transformation for the automotive industry. “Smartphone on wheels” – that’s a phrase I’ve been hearing since 2015. The smartphones have already been software-defined for more than a decade and the autos journey has just started towards the Software Defined Vehicle “SDV”.
The need for SDV
In simple terms, SDV is needed to match the customer requirement by removing the inherent complexity and making the vehicle more innovative, customer-centric, and personalized while remaining profitable and creating room for collaboration.
What is the SDV and its benefits?
What will the next generation of vehicles look like? What can I expect as a consumer? Just like smartphones are getting new features throughout their lifecycle, SDVs are also supposed to get new features and updates over their lifecycle of 10-15 years. Consumers can expect and enjoy the new experiences driven by software, it shall bring enhanced performance and improve security with time, it shall also bring a much more personalized and customized experience for the user/driver of an SDV.
What are the industry challenges?
Now, with this context of SDV, all the global OEMs have been investing to develop Software Defined Vehicles. This requires radical changes and modifications in the vehicle architecture as well as adding many digital services to and around the vehicle becomes a must. A different perspective is to see the vehicle as a digital device, where the vehicle itself is seen as hardware and all the software needs to be added and managed for a longer lifetime than a smartphone. There are challenges in this simple vision of the automotive industry developing SDVs.
SDV is not just a technical challenge, it is more than that. We see the following key challenges:
- Shifting from hardware-defined to software-defined product development
- Establishing the right organization, structure and processes
- Enabling and establishing the right technical skills
- Establishing the right collaboration with partners/suppliers
- Speed of adaptation and implementation
Today, vehicle architecture is too complex and costly for OEMs to deal with.
The vehicle architecture must be redefined. Typically due to its overall complexity, it takes 4-8 years to develop a new architecture before vehicles can be produced and launched in the market. The vehicle software is tightly coupled with the ECUs and vehicle networks. There is no software release, without underlining hardware release. But there is also no software reuse. With every vehicle project, the vehicle software is written from scratch. In many cases, different suppliers, don’t know about the preview projects. Faster product development cycles are required to meet the market demand.
Automotive cloud becomes business critical for OEMs
Along with the new vehicle architecture, the automaker must develop an automotive cloud that is seamlessly connected to the vehicle to provide security, updates, external capabilities, and the digital twin. Along with the in-vehicle architecture, the automotive cloud is another critical component in the development of next-generation vehicles. As the automotive cloud manages the complete lifecycle of a vehicle including service and maintenance, the external world via APIs, the providers of content such as live traffic data, road condition data, streaming content, payment providers for tolls, and parking, to name a few.
The cloud infrastructure is the foundation for all business logic and vehicle-related computes. With our 25 years of experience in automotive technology and automotive cloud, we can share the insights that most car manufacturers do not achieve the scale of economy and cost advantage with such a connected car backend, as the cloud infrastructure and data storage costs explode with the number of vehicles onboarded to it and OEMs are required to run the vehicles for multiple years, so any mistake or wrong decision made with the automotive cloud can cost them quite a lot. We also have seen that even the best software developers struggle with such an automotive cloud delivering the required features in time. In this industry, many SOPs fail due to such mistake of believing that just by hiring the best software talent or software company, they would solve the problem. The key here is to have the right skills combination such as vehicle-software engineers who are developing intelligent vehicles, the software-defined vehicles, together with the engineers of the OEMs – vehicle and automotive cloud development at the same time.
The road to SDV is a long one, and several industry experts need to work together to reach the milestones. The federate initiative, supported by the European Union, is also working on SDV together with the major automotive players along with open-source organizations like Eclipse SDV and COVESA. An interesting publication is their vision and roadmap – see the diagram above.
Evolution of in-vehicle architecture: On Board
There is a lot of work to be done on the vehicle side. Today’s vehicles have hundreds of ECUs, so a lot of money, time, and effort is spent just integrating and validating them, rather than producing new features for the end customer. High Computing Platforms “HCPs” provide the computing power needed to deliver the complex features expected in every new vehicle today. As a first step, a so-called domain architecture has been introduced, with dedicated HCPs for the major domains such as powertrain, body, ADAS, IVI and navigation, telemetry, and so on. The current challenge for in-vehicle architectures is the design of so-called zonal architectures. The main idea is to provide a high computing core that can be provided by a large HCP or a cluster of HCPs. This core is surrounded by so-called zonal gateways, which are connected via high-performance, low-latency/time-sensitive networking (TSN) Ethernet. The classical automotive ECUs, sensors and actuators are connected to the corresponding zonal gateways via classical networks. All communication can be managed centrally by the HCP core instead of multiple ECUs having their communication paths with different buses (LIN, CAN, Flex Ray, MOST, etc..). This approach dramatically reduces the amount of wiring harness required, which in turn saves on material costs and simplifies production. On the other hand, zonal architecture invites most of the vehicle software components to run on the high computing core. For this purpose, a global and powerful vehicle abstraction layer is introduced, which reduces the tight hardware dependency.
A difficult challenge for zonal architectures is to decide which software modules should remain on the ECU hardware. In specific safety-related cases where the lowest latency is required, such as brake activation, the most reliable approach will be to keep the brake software on the ECU. Another potential of zonal vehicle architectures is to reduce the number of ECUs required in the vehicle, especially those for non-safety functions. In other words, the onboard evolution aims to simplify and improve in-vehicle communication for data-intensive operations required for AD/ADAS, safety, monitoring, and digital twin, to name a few.
Standardization is urgently needed to open the door for others to collaborate. Through hardware abstraction, virtualization, and the provision of well-documented high-level APIs, OEMs should create the space for faster innovation in the areas of smart cockpit, multimedia/infotainment, AD/ADAS, and powertrain-related functions to enhance the in-vehicle experience with the combination of the outside world.
Evolution of Off-Board / Automotive Cloud
To take full advantage of the possibilities offered by the outside world, be it the world of OEMs, the marketplace, the Internet or the World Wide Web in general, the connected vehicle backend or off-board components need to be adapted to SDV. In particular, the automotive cloud is responsible for the following key areas:
- Manage in-vehicle software through over-the-air (OTA) updates
- Management of in-car applications through a marketplace inspired by the Google Play Store or Apple App Store
- Management of content providers such as navigation providers, charging providers, etc.
- A digital twin that provides a detailed, safe and secure vehicle abstraction in the backend, allowing communication with specific parts of the vehicle and provisioning of changes to deliver innovative vehicle enhancements and features to the customer.
- Transform telemetry data into products through data analytics and data-driven services and offerings. In addition, various data marketplaces can be integrated for data monetization.
Tools / Collaboration Methods / How-To
The automotive industry has already started the transition to embrace the agile way of working to get the best result at the best time. The other is to break down departmental silos and have cross-functional product teams responsible for the end-to-end product, something that big tech giants like Google and Apple have practised in their organizations. Automakers need to change their old habit of thinking that they just build a car, and the software comes after the car. So, all the stakeholders from the product portfolio, E/E and automotive engineers, quality, digital experience, vehicle software engineers, and aftersales need to come together to define and develop the product, to achieve the milestones of the evolutionary journey of SDV.
Because of the limitations and constraints, we will stop here. More details on tools and collaboration methods will be written in our next articles.
Daniel Elhs
Business Director
and Subject Matter Expert
SDV, Valtech Mobility GmbH
Nilesh Chauhan
VP Automotive and Mobility, APAC
and MENA Region
Valtech Mobility GmbH
Disclaimer: The views expressed by the authors are their own and do not necessarily reflect the views of FMM magazine.