What Is Cloud-Native in Automotive?

Cloud-native is an approach to cloud computing that is essential to expanding functionality and streamlining software development and deployment in automotive. Cloud-native vehicles will become intelligent edge devices, not only evolving through software updates delivered via the cloud but also leveraging the vast processing power available in the cloud for specific functions.

However, to optimize the benefits of a cloud-native approach, vehicles must be designed from the beginning with cloud-native in mind. Some features and services will always have to run on the vehicle — such as safety-critical applications ­— which means that the vehicle’s hardware and software architecture must be engineered to balance processing between onboard compute and the cloud.

But what are the benefits of cloud computing that make rearchitecting the vehicle worth the effort? In other words, why does the automotive industry have its head in the cloud?

What is driving cloud computing?

In the automotive industry, vehicles are evolving beyond mere modes of transportation and are becoming software-defined devices on wheels. Consumer demand for increased software functionality has made a cloud-native approach essential.

Compute power is expensive, and there are already more than enough features competing for limited onboard processing power. Cloud connectivity offers a powerful computing resource to enable premium user-experience features, reduce costs or both by offloading processing power to the cloud, thus enabling OEMs to reduce vehicle hardware costs while improving performance. Finally, there are some features or services that are possible only by aggregating inputs from other devices — including a fleet of vehicles — which naturally favors the cloud for consolidation, analytics and input back to the vehicle.

Will vehicles run entirely in the cloud?

Cloud-native principles have already revolutionized various industries, including web development and IT. Consumer electronics have shifted more of their processing to the cloud as well. Most mobile device applications are designed to minimize their footprint and rely on cloud processing to perform their primary functions. However, this level of cloud interconnectivity is not appropriate for many automotive applications. As a result, vehicles will not run entirely in the cloud anytime soon.

Low latency and cybersecurity are essential for safety-critical systems, so core vehicle functions must be able to completely rely on local high-performance compute resources. But the cloud can still enable OEMs to expand functionality over time and generate new revenue streams, and 5G connectivity opens up new opportunities for processing at the edge, such as low-latency processing at local base stations.

By leveraging the cloud for additional processing tasks, vehicles can run more powerful algorithms for long-term vehicle and behavioral data analysis. This will enable them to deliver incremental feature improvements for end users, enable predictive maintenance, optimize performance and reduce downtime. Through the digital feedback loop, cloud-native solutions support scalable data processing, allowing for historical trend analysis and strategic decision-making based on comprehensive, aggregated datasets at the fleet level to improve software development and automation.

Benefits of cloud-native in the automotive industry

Cloud-native development offers immense potential to revolutionize the automotive industry in various domains. Here are some of the other benefits of a cloud-native approach:

End-user personalization: A cloud-native approach makes it easier for OEMs to introduce new features through over-the-air updates, allowing them to generate new revenue streams and enabling vehicles to feel like new to customers every day. OEMs could even use vehicle data to offer customized features such as a battery management software designed around a user’s specific driving patterns.

IoT connectivity and V2X: Cloud-native development enables internet of things (IoT) communication — improving the safety, comfort and convenience of vehicles by allowing them to communicate with a wide range of devices. For example, support for vehicle-to-infrastructure communication through 5G enables a vehicle to communicate directly with traffic lights to find out exactly when a red light will turn green, which lets it save energy and time by maintaining a steady speed as it approaches. Vehicle-to-grid is another example of vehicle-to-everything that lets EVs communicate with the public power grid; this gives OEMs greater control over the charging process so they can optimize charging cycles and extend the overall life of the battery. Software-as-a-service could enable OEMs to easily employ each of these use cases by subscribing to an independently managed application via an API — with software updates, bug fixes and general software maintenance being handled by the SaaS provider.

 Scaling compute: Cloud computing allows developers to dynamically scale compute resources based on demand. Packaging applications and their dependencies into software containers makes them easy to deploy across different environments, and utilizing services gives developers the flexibility to make changes to independent software functionalities without affecting adjacent services.

 Shared mobility: Down the line, a cloud-native infrastructure will be necessary to enable shared mobility. By providing a scalable infrastructure for processing vast amounts of real-time data, a cloud-native architecture will support ride-hailing, vehicle tracking and route optimization, allowing seamless coordination between users and services.

Cloud-native requires an end-to-end approach

 Executing a cloud-native vehicle architecture comes with unique requirements. Cloud-native vehicles need a hardware architecture that is independent of the software that runs on it, and the ability to make targeted updates to safety-critical systems.

 A microservices architecture helps achieve this by breaking complex monolithic software applications into smaller, more manageable components that are easier for developers to update and maintain. This type of modular approach enables greater scalability, resilience and agility compared with traditional, monolithic applications.

 A DevSecOps platform that enables continuous integration, continuous deployment and continuous testing is essential to managing the lifecycle of the vehicle during development and after the point of sale.

 As the only provider of an integrated vehicle brain and nervous system, Aptiv is well equipped to help OEMs adopt a cloud-native approach. We are integrating our Smart Vehicle Architecture™ technologies with Wind River Studio, a cloud-native toolset for developing, deploying, operating and servicing mission-critical intelligent systems across the edge.