VERSION V1.0
As part of our continued community engagement, the following list of principles serve to provide an understanding of how the Eclipse Software Defined Vehicle TAC (Technical Advisory Committee) intends to work with projects, effectively serving as the guiding star for the TAC. It also provides insight into what the TAC is looking for in terms of continuity and evolution over time. Since the SDV's inception we have endeavoured to capture the lessons learned and iterate on these principles to serve the working group, the community, the adopters, and the projects.
- We are code first.
- Projects are self-governing to enable community driven success.
- Aim for success by selecting robust, sustainable, well engineered projects.
- No kingmakers. Similar or competitive projects are not excluded for reasons of overlap or competition.
- Don’t reinvent the wheel. Adopt and promote useful work happening in other industry groups.
- No single stack. Encourage interoperability for the emergence of a variety of stacks and distributions to serve the community and adopters.
- Define specifications derived from open source projects which have demonstrated broad adoption and real world use.
- We want to build automotive grade software (from QM to ASIL-D) which will be used in series production.
- We are here to help our projects.
The SDV community is first and foremost focused on shipping working open source software intended for adoption by the automotive industry.
The SDV steering committee, upon advice from the TAC will identify Eclipse Foundation open source projects which are within the purview of the working group. The purpose of the working group is to complement, but in no way replace, the self-governing project community who are solely responsible for driving the technology forward. As per the Eclipse Development Process,
all Eclipse Foundation projects and their Project Management Committees are responsible for all technical decisions within the project.
We are looking for high-quality, high-velocity, broadly adopted, open source projects that further the goals of providing automotive software platforms for tools, in-car, edge, and cloud. The SDV is a badge of quality. velocity, and adoption. SDV projects should be on a path to being tools that users can trust, broadly work together, and that meet other automotive industry quality criteria.
Identify projects that have a real shot at being a useful tool in the evolving portfolio of open source automotive technology. This will require a mix of mature and early-stage projects.
Some project attributes for consideration by the TAC:
- Has open, transparent, and vendor-neutral technical governance, making it more likely to successfully graduate from incubation in a timely manner.
- Has users, preferably in production; is a high quality, high-velocity project (for incubation and graduated projects).
- Has a committed and excited team that appears to understand the challenges ahead and wishes to meet them
- Has a fundamentally sound design without obvious critical compromises that will inhibit potential widespread adoption
- Has an affinity for how SDV wants to operate
Simply put, the TAC will never deny a project’s inclusion in SDV because there already exists similar, competitive, or overlapping projects.
- Many problems in technology have more than one solution. Markets are good at finding them. There may be multiple good OSS projects for solving a problem that are optimized for different corner cases.
- Often multiple solutions get widespread use, for example, because they are optimized for different constraints. We don’t want to get in the way of that by insisting that one technology is “the answer” for each functional gap that we can identify today. We believe that market adoption and the user community provide a good mechanism for pushing the most appropriate projects forward over time. We want projects to enjoy the support of the SDV during that process.
- There is no “one true stack”. Automotive systems and applications cover many different use cases with different needs. Many architectures are reasonable. And there are many scales from one node to many, from low to high latency, etc. So “one size does not fit all”.
The TAC recommends projects with a real chance of achieving widespread adoption, but it does not pick a single winner in each category. Similarly, the TAC recognizes that projects take time to grow, mature, and attract interest, and is welcoming to new projects as they work to drive participation and adoption.
The automotive industry has numerous other initiatives that add value to the ecosystem. SDV is an initiative of initiatives and wherever possible will adopt existing standards, specifications, and open source implementations that help further our mission. The TAC will play an advisory role to the open source projects in determining which initiatives to align with.
SDV will deliver multiple, competing stacks and distributions for adoption by industry. The TAC does not pick a “winning stack” - i.e., vertically integrated set of projects as a solution for multiple application problems. This means that SDV is not promoting a single, monolithic stack of technologies.
As SDV open source projects grow and mature, and as SDV’s distributions and stacks evolve, interoperability, stability, and sustainability will become as important as a high rate of innovation. The SDV TAC will work with the SDV Specifications Committee to identify which projects may provide inspiration for future SDV specifications. The projects most likely to be selected will be those which have demonstrated through real world adoption that the industry would benefit from the added stability and opportunity for independent implementations afforded by a formal specification.
As per the Eclipse Foundation Specification Process, SDV specifications are intended to facilitate the development of independent implementations. Each specification will provide a specification document, a compatibility test suite, and at least one open source compatible implementation.
The goal of SDV is to provide open source technologies that will be broadly adopted by the automotive industry for real world use. To that end, SDV will strive to deliver automotive grade software (from QM to ASIL-D) which will be used in series production. The TAC will play a role in defining additional processes, consistent with the Eclipse Development Process, which SDV projects may choose to adopt in order to consistently implement attributes such as quality management, functional safety, and supply chain security. The TAC will also define processes, maturity models, and requirements which SDV-related projects may choose to implement in order to participate in SDV branded distributions and releases.
We want to be able to say that SDV is a net positive for big & small projects. Doing so requires coordination with project leadership.
Project needs may include test automation and CI, cloud resources to test on, clear documentation, per-project marketing & evangelism, roadmaps for interop, and advice from experts on governance, licensing and scalability. And we need to make sure project contributors see what value they are getting & are not afraid to ask for help!
Note: These principles were presented at the first SDV WG TAC meeting on 27/04/2023.