The automotive industry is moving through an unprecedented time that may prove to be as critical as the era when motorized vehicles first took the place of a horse and carriage. There is a feeling of anything being possible as the industry and consumers move away from gas-powered vehicles to all-electric ones.
The U.S. government signed an executive order to end purchases of gas-powered vehicles by 2035, in a move to lower emissions and promote electric cars. Simultaneously, global automakers plan to spend more than half-a-trillion dollars on electric vehicles and batteries through 2030, amping up investments aimed at moving car buyers away from fossil fuels and meeting decarbonization targets. The move toward zero carbon emissions is prompting automakers to align their vehicle roadmap and implement the necessary technologies to transition toward electrification and autonomous driving.
The latest SAE International J3016 standard on “Levels of Automated Driving” has encouraged automated vehicle implementers to explore the space between Levels 2 and 3. With the latest advancements in sensing and processing technologies, OEMs recognize the need to enhance their current L2 ADAS systems to improve overall passenger safety and the driving experience.
According to ABI Research, autonomous passenger vehicle shipments are expected to reach over 21 million units by 2030. The most significant increase will be robotaxis, with a 91% compound annual growth rate (CAGR), and Level 2+ semi-autonomous vehicles, with a 34% CAGR.
The more a driver is removed from the loop under any condition, such as L3 systems and beyond, the more complicated and costly the advanced driver-assistance systems (ADAS) become. So as the move toward full autonomy continues, L2+ systems offer OEMs a solid middle ground from which to introduce new features within the scope of today’s regulatory standards — without additional liability and cost burdens that will eventually come with higher levels of autonomy. This allows OEMs to improve their ADAS system performance and reliability by integrating the latest camera, radar, LiDAR, and fusion technologies while evaluating driver-monitoring systems for more robust, hands-free driving.
Connectivity brings new monetization opportunity
Analogous to ADAS progression, the automotive industry is witnessing major transformations within the interior space of modern vehicles. To keep pace with their industry peers, OEMs are seeking to implement the latest and greatest features, such as automated park assist, 3D surround view, augmented-reality heads-up display, in-cabin monitoring, 5G connectivity, vehicle-to-everything (V2X), and more. Connectivity plays a central role for accommodating the ever-broadening range of user preferences.
Until recently, the potential for the connected vehicle has been somewhat constrained due to limited capabilities within the cellular network in terms of latency, throughput, and data processing.
Both automakers and telecom companies are aware of the broad-based market potential that tomorrow’s connected car can bring. The two industries are teaming up on working toward solutions to address potential constraints. Continuous rollout of 5G, adaption of high-speed wireless connectivity (such as Wi-Fi 6), and transition from a centralized cloud-compute scenario toward a more distributed edge-cloud–computing scheme are among the various ways to improve vehicle-to-cloud robustness and performance.
Applications requiring low latency (such as V2X) and increased data analytics (such as HD mapping) are becoming safer and easier to implement. Thanks in part to the innovations at Tesla, OEMs are realizing the benefits that over-the-air (OTA) firmware updates can bring to their future business models, whether to enhance vehicle performance or enable new features and services to existing platforms. There is a variety of new use cases allowing data-monetization opportunities for carmakers following the original point of sale.
Advancements in technologies are reshaping the automotive ecosystem and ways for which companies develop their future mobility business models. To remain competitive in an evolving market, automakers must transition from a distributive ECU architecture approach toward those that are software-defined in order to allow greater flexibility for expansion enabled by domain- and zonal-type architectures.
In domain-based architectures, the vehicle’s logical or software systems are divided into several major domains or functional groups, including ADAS, body control, cockpit, engine, and chassis. Each domain has a central domain controller for processing and controlling data flow. For example, a cockpit domain controller may consolidate many features and functions into one or more large processors for managing multi-display features such as infotainment, cluster, heads-up display, surround view, driver, occupant monitoring, and others.
Domain architecture features:
- More flexible and scalable hardware/software environment
- Central gateway, which supports safe and secure OTA
In a zonal architecture, the vehicle is partitioned into various zones with each zone, such as high-speed Ethernet, having a controller connected to a centralized vehicle controller via a backbone. Each zone controller manages the network data deployment, power, and other information required by each end node, including cameras, radar sensors, displays, etc., establishing a streamlined in-vehicle network.
Zonal architecture features:
- Highest flexibility in terms of hardware/software deployment
- Optimal cable and wire harnessing
Both domain and zonal architectures enable greater processing power, efficiency, and throughput, resulting in a highly scalable, cost-effective platform for which auto manufacturers can deploy new features while enhancing vehicle performance and overall customer driving experience.
At the heart of these software-defined architectures are high-performance systems-on-chip (SoCs) consisting of a comprehensive set of IP building blocks, including multi-core CPU and GPU subsystems, AI accelerators, ADCs and DACs, and others. Socionext’s unique Solution SoC business model provides a scalable and efficient means for meeting the demands for performance, functionality, reliability, and safety.
Socionext’s team of dedicated system architecture engineers is equipped with decades of proven experience and expertise in developing and implementing highly complex SoCs.
Socionext collaborates with partners worldwide to provide and propose platform and core technologies ahead of the curve, delivering complete solutions from system design to production and quality control. The company actively participates in the customer’s development process from the early stages and provides them with the needed support to help differentiate their products according to their specific needs.