"Cloud technology will never be used in automotive"? When discussing the application of cloud deployment in the automotive industry, it was once widely believed that cloud technology was unsuitable for product development in this sector. Practical challenges such as intellectual property protection, data security, lack of transparency in processes, high migration costs, and the conservatism of IT departments served as significant barriers to the adoption of cloud technology in automotive development. However, with the success of electric vehicle giants like Tesla and other companies, the industry's development trends have become unmistakable: the automotive sector is moving toward a more software-driven future. An increasing number of OEMs are repositioning themselves, shifting their focus to software and technology. Several factors are driving this transformation, including advanced driver assistance systems (ADAS), e-mobility technologies, and the development of infotainment and comfort systems. Overall, the automotive industry has begun to adopt cloud technology to accelerate the design of automotive software, pushing it toward becoming more advanced and intelligent.
This article focuses on exploring the advantages and considerations of applying cloud technology in the automotive industry, as well as its practical application in ensuring the quality of automotive software design.
Cloud Deployment of MBD Toolchain
Key Elements of Cloud Technology
Cloud technology is a computing model that provides on-demand access to computing, storage, networks, and other IT resources. In this model, computing power and applications can operate in centralized server farms located anywhere, unconstrained by physical space. These servers may be deployed locally by OEMs or hosted on public cloud servers in various regions. A critical resource in cloud technology is the virtual machine (VM), which serves as a virtual computing environment. Within these servers, VMs are used to handle computations and can be dynamically deployed or undeployed based on workload demands. In theory, cloud technology offers virtually unlimited scalability in computing power. Personal computers also play a vital role, serving as the primary interface for interacting with the computing capabilities of cloud data centers. Bandwidth resources are equally significant in cloud computing. For cloud applications to function effectively, sufficient bandwidth is required to enable devices to establish fast internet connections. Additionally, it is essential to monitor the actual usage of application services to ensure smooth and reliable operation. Cloud technology enhances development flexibility, simplifies scaling, and optimizes resource utilization. From a business perspective, it reduces costs, improves efficiency, and supports agile workflows.
There are several key concepts associated with cloud technology. Container is a technology that packages applications and their dependencies together, enabling seamless deployment, execution, and management of applications in a cloud environment. One common implementation of container technology is Docker, an open-source containerization platform. Docker helps developers package applications and their dependencies into containers, allowing them to run in any environment that supports Docker, eliminating issues caused by environmental differences. For clustering technology, Kubernetes (commonly referred to as k8s) is a container orchestration system that helps enterprises manage containerized applications efficiently. Automation, as a core concept in cloud computing, aims to reduce costs and improve efficiency. Tools like Jenkins, Bamboo, and Git are frequently mentioned in the context of automation pipelines and build pipelines. Jenkins is an open-source continuous integration tool written in Java, similar to Bamboo. Jenkins supports the automation of building, testing, and deployment processes, improving the efficiency and quality of software delivery. In addition, version control systems, such as Git, are also essential. Git allows teams to track project changes and collaborate effectively on development.
A repository is used to store software elements related to cloud technology. The primary types of cloud services are IaaS (Infrastructure as a Service), PaaS (Platform as a Service), and SaaS (Software as a Service):
- IaaS provides computing, storage, and network resources, allowing users to pay based on actual usage.
- PaaS offers platforms for developing, deploying, and operating applications.
- SaaS delivers pre-built software applications.
Other service models include:
- CaaS (Container as a Service), which focuses on providing foundational network resources and operating system services to consumers.
- DBaaS (Database as a Service), which gives customers access to databases without the need to deploy or manage the underlying infrastructure.
- Serverless computing, also known as FaaS (Function as a Service), is based on PaaS. It provides a micro-architecture where customers do not need to deploy, configure, or manage servers. The cloud platform takes care of all server-related services required for running code.
For small companies, cloud computing enables them to host websites or applications without investing in their own hardware and software, thus reducing costs. For large enterprises, cloud computing allows the expansion of IT infrastructure to meet diverse business needs efficiently.
Challenges & Opportunities in the Industry
Scalability is the foremost and critical advantage of cloud computing. Flexible tool configurations enable cloud computing to better adapt to various workloads and work scenarios. Centralized management allows for real-time monitoring, ensuring transparent and controllable operational processes. Global standardization is equally important. Cloud computing provides a streamlined workflow, and its ease of maintenance ensures system reliability and efficient utilization. These advantages help enterprises reduce costs while improving transparency.
There is no doubt that cloud technology also faces several challenges. Firstly, the application of cloud computing may require users to alter their original tool usage patterns, potentially leading to difficulties in adaptation. Users' acceptance of these changes becomes a significant challenge. Additionally, migrating the entire project to the cloud can be time-consuming, potentially slowing down project progress or execution. Enterprises need to conduct thorough preparations before migration to address these issues effectively. Another challenge is the incompleteness of cloud-based toolchains. Some critical tools may not yet be fully adapted to cloud environments, which can create obstacles. Continuous technological development is necessary to enable a fully cloud-integrated toolchain. Finally, adopting cloud technology may increase the enterprise's dependence on third-party services, which introduces certain risks.
Use Cases of Cloud Deployment in the Automotive Industry
In the development and testing of automotive software, cloud deployment architectures are primarily applied in two forms: Desktop Cloud and Process Cloud. Desktop Cloud is user-centric. In this case, the end user, typically someone working at a PC, triggers applications through a conventional PC, while computing power and data are stored in server farms in the cloud. From the user experience perspective, this approach is almost identical to working on a conventional PC. The software and applications are used interactively. Process Cloud, on the other hand, is process-centric. At the user end, automated workflows trigger applications, relying on the automated pipelines introduced earlier, such as Jenkins, Git, and Bamboo. The results are automatically distributed in the form of reports or events.
A specific architecture case can be analyzed as follows. For software development, when developers and testers commit a model update or test change to a Git repository, this action proactively triggers an automated pipeline on the Jenkins master node. This pipeline contains a series of descriptions for the test environment. When the test environment description files are modified, the Jenkins master node, acting as the central hub of the data center, is automatically triggered to control and distribute these changes in the pipeline configuration. The Jenkins master node first requests deployment from the agent nodes. Then, the agent nodes request the download of the corresponding Docker containers from the cloud repository.
As shown in Figure 1, the Docker repository contains pre-deployed container packages tailored for different platforms (Windows or Linux). The container package that meets the system requirements is then requested and deployed to the Jenkins agent node. The container package (Docker) receives commands from the automated pipeline and subsequently checks out the necessary information, such as models and configuration files, from the Git repository. It then performs static analysis based on the specific instructions. The Jenkins master node then requests the results of this analysis, which are sent back and automatically stored. Finally, the Jenkins master node initiates a request to the agent nodes to undeploy the containers (Docker) that are no longer needed. It also generates and sends the final report to the developers or testers.
From the user’s perspective, user clients can also achieve scalability, utilizing terminal servers to complete the entire process of development, testing, and analysis (as illustrated in Figure 2). Users, whether individual or multiple developers or testers, can log in through the terminal server. The server matches and configures different toolchains based on varying work environments and requirements. If specific tool licenses are required, requests will be sent to a floating license server, with licenses being hosted in various locations as needed. To ensure security, servers are typically configured in redundant or geographically redundant modes. In this setup, the commit operation automatically triggers the Jenkins controller which distributes the corresponding automation pipeline to the agent nodes. Using Kubernetes these agent nodes can be dynamically scaled, allowing for a seamless transition from fixed capacity to flexible, on-demand resource allocation, effectively expanding the cloud environment.
For streamlined process support in cloud architecture, MES offers preconfigured container templates to simplify the setup and configuration of software tools within containers.
License Model & Cloud Deployment of Toolchains
Referring to the keywords introduced previously, cloud service models can be broadly categorized based on the delivery mode of the cloud service provider:：
- Software as a Service (SaaS)
- Platform or Infrastructure as a Service (PaaS or IaaS)
In Software as a Service (SaaS), user applications are integrated into the hosted service, and the business model primarily operates on a pay-as-you-use basis. In Platform or Infrastructure as a Service (PaaS or IaaS), attention is directed toward the computing power and operating system capacity leased by the customer. Platform as a Service (PaaS) includes operating systems and databases. Infrastructure as a Service (IaaS) includes hardware and virtual machines (VMs). Customers using PaaS or IaaS retain ownership and licensing of their application software. They can define usage costs through customized plans. Under high workloads, additional capacity becomes critical, making flexible dynamic subscription models an essential solution. These models provide additional support when usage exceeds regular limits.
For the MBD toolchain, the industry standard primarily involves hosting infrastructure. Applications and toolchains are often stored within the company, as this architecture facilitates monitoring and flexible control over the development process. For automotive manufacturers, Infrastructure as a Service (IaaS) is the primary method of applying cloud technology. There is a growing demand for dynamic licensing models across the industry, and MES aligns with this trend by offering flexible licensing solutions.
The MES toolchains are fully adapted to cloud-based licensing models and are available in both Windows and Linux versions. For the automotive industry, subscription-based licensing is the mainstream choice. One licensing model is the Developer Desktop Cloud, designed for interactive cloud environments and tailored for migrating desktops to the cloud. For licensing coverage, the Automation Cloud Global enables fully automated and highly dynamic global licensing, accessible from anywhere in the world, designed specifically for migrating processes to the cloud. Additionally, MES currently supports setting up the MBD toolchains in the cloud. MES provides cloud-based sandbox environments for pilot projects and configuration for Docker containers. Furthermore, MES also supports automation pipelines configured in the cloud.
The Developer Desktop Cloud licensing model is designed for interactive cloud environments, facilitating the migration of desktop-based development to the cloud. For fully automated and highly dynamic global licensing, the Automation Cloud Global model enables seamless access from anywhere in the world, specifically tailored for cloud-based process migration.
In addition, MES supports the setup of MBD toolchains in the cloud, offering cloud-based sandbox environments for pilot projects, Docker container configuration, and automation pipeline deployment.
As illustrated in Figure 3, MES provides end-to-end support for cloud deployment—from technical consulting to tools and workflows, and ultimately to build and migration—helping clients successfully transition their toolchains to the cloud.
More Offers from MES
Webinar: Let Your Toolchain Fly: MBD in the Cloud
OEMs and Tier 1 suppliers have now discovered the benefits of cloud technologies and started migration of key development toolchains to the cloud. In this webinar, we showcased how leading OEMs and suppliers push their software development tools into the cloud.
Find Your Ideal Solution: Flexible MES Licensing Options
Want to learn more about MES tools and services? Check out the introduction to our flexible licensing options!
