Cars are morphing into upgrading, high performance computing devices. From the ability to connect and stream music, with in-built map functionality some 10 years ago, cars have accelerated ahead to transform into computers on wheel as both hardware and software can do much more than tell you which turn to take. Examples of new technologies include cellular connectivity, cloud management, data access, and data analytics. This will be driven by the maturity of 5G capability, developments with vehicle-to-everything communication (V2X), and the advent of autonomous vehicles.
The latest development in computing technologies are cars was announced by NVIDIA and Mercedes-Benz that they are collaborating on a next-generation architecture for Mercedes-Benz vehicles that will enable over-the-air (OTA) upgradeability and automated driving functions. Starting in 2024, every next-generation Mercedes-Benz vehicle will include this first-of-its-kind software-defined computing architecture that includes the most powerful computer, system software and applications for consumers, marking the turning point of traditional vehicles transforming into computing devices.
Nvidia’s new DRIVE AGX Orin platform will be the centerpiece of the architecture, integrating inputs from multiple sensor types distributed throughout the vehicles, as well as utilizing the entire NVIDIA DRIVE software stack from the OS through perception, planning and mapping.
While major automakers have taken deliberate steps toward upgradeability and partial/full autonomy, there is a growing number of ‘new challenger’ auto OEMs seeking to contest the current leaders with similar capabilities – and potentially earlier than 2024.
The latest partnership is expected to result in multiple new capabilities for future Mercedes-Benz passenger vehicles. First among these is the ability for OTA upgradeability of the vehicles, which Mercedes-Benz says will be the first time that it will have the ability for OTA updates to alter the driving function itself. Current cellular gateways and connectivity are primarily limited to infotainment applications such as mapping and traffic information but do not extend to the driving capabilities of the vehicle.
The automaker stressed its goal to leverage OTA to enable customers to purchase new features and upgrades to the vehicle throughout its ownership period and allow the OEM to continually upgrade safety and convenience features. Second among these capabilities is new automated driving features, including Society of Automotive Engineers Level 2 and 3 driver assistance such as automatic emergency braking and adaptive cruise control, and Level 4 ADAS like address-to-address autopilot and valet parking. Mercedes Benz notes that these vehicles will be sold globally and are not dependent on future availability of 5G networks.
The announcement from NVIDIA and Mercedes-Benz will result in a ‘software-defined car’ that can be upgraded continually with improvements and features. This is a stated goal of the major automakers, although OEMs have been guarded about elaborating on architectures within their vehicles and instead have forecasted ever-evolving timelines for semi-autonomous and fully autonomous driving. Ford is still targeting 2021 for the debut of its self-driving car fleet. GM once thought it would be possible to launch an autonomous taxi service by the end of 2019, but it has had to push back that date and has not specified a new timeline.
Toyota is targeting 2020 for self-driving on highways but has not announced when it expects to achieve Level 4/5. Honda (via Acura) is targeting 2020 for the release of a vehicle with Level 3, and 2025 for Level 4. Hyundai’s CEO has said that he thinks it is a reasonable goal to have something ready for market by 2022. Renault-Nissan is pushing for ‘some autonomy’ in urban conditions by 2020, and full autonomy by 2025. BMW is projecting Level 3 automation by 2021, with one senior executive indicating that it might be possible to achieve Level 4 and 5 capacity by that point. Volkswagen has been unwilling to give a firm timeline, but recently invested $2.6bn in Ford’s partner, Argo AI.
The most vocal of the ‘new challenger’ OEMs is Tesla, which is often synonymous with autopilot and autonomy, and currently has a fleet of just over one million vehicles globally. These vehicles are on their third generation of silicon and are already upgradeable over the air with changes to driving functions. The most recent computing architecture shipping in new models today utilizes internally developed silicon, in contrast to Tesla’s second-generation NVIDIA-based architecture and first-generation Mobileye (now Intel)-based technology. Tesla delivered 367,500 vehicles in 2019, compared with Mercedes-Benz’s 357,729 during the same period.