How can you develop and document software code for aerospace engineering?

Know the purpose of your code. What will it be used for? Who will use it? How long will it be used for? Will it ever be used to present data to customers?

Software requirements are multifaceted and play a pivotal role in software development. They are a means of understanding user needs, aligning with business goals, guiding development efforts, ensuring quality, and facilitating communication among stakeholders. A holistic approach to software requirements considers all these perspectives to deliver successful software projects.

the waterfall approach can be suitable for aerospace engineering due to well-defined regulations and stable requirements. The phased structure ensures thorough verification at each stage. Nonetheless, rigidly defining the end product at the project's outset can be futile in this dynamic industry. We should maintain flexibility by incorporating agile elements like incremental planning and adaptability to effectively balance predictability and responsiveness.

Requirements Engineering is a very important process to ensure code safety, quality etc. I recommend to use standards, such as ISO/IEC/IEEE 12207 and others for software requirements generation and management.

Plant UML is a great tool for this as its format is the easiest I have used for tracking changes in version control and merging changes when working collaboratively.

Software design is a crucial phase in the software development process. It involves creating a plan or blueprint for how a software system will be structured, organized, and implemented to meet the specified requirements. Effective software design lays the foundation for a robust, maintainable, and efficient software application. Here are some key aspects and principles of software design: Modularity Abstraction Encapsulation Separation of Concerns Scalability Flexibility and Extensibility Simplicity Testing and Debugging User Experience (UX) Design Documentation Security Performance

In a case such as aerospace engineering where the software is this mission critical, it is important to always consider Murphy's Law; "Anything that can go wrong, Will go wrong!" possibly at the worst time and in the worst way. It is therefore crucial to design each part of the system, software included, to be as robust as possible with fail safes, recovery tools and backups in place. While this should be the practice in the whole development process, it starts from design. It is also the most fundamental part of the process. A flawed piece of code might break a part of the system but a flawed design breaks whole of it.

In addition to fulfilling requirements on customers' needs, system and engineering features, regulatory obligations and etc, future software must be extensible. Requirements for complex systems almost always are not constant and evolve at software lifespan. Taking into account specific domain knowledge and choosing appropriate technical solutions makes it easy to design the right architecture.

Each programming language has its own features making some kinds of tasks easy. But stack fragmentation is a costly thing so frameworks/programming languages diversity minimization at a project is a good idea.

Choose the best software coding language for the job. While each language has its own advantages, each one is best for a specific type of job. Determine if your code is primarily for: visualization, data processing, image analysis, etc.

The code should be self-documented. It is essential to use readable and self-explanatory variable names and method names. By choosing descriptive names, others can easily understand the purpose and functionality of each element within the code. The method names should have at least one verb. The code should follow the coding standards chosen by the language community. Having a comprehensive test suite is beneficial. Tests not only ensure the correctness of the software but also serve as a form of documentation by demonstrating how the software can be used. Provide minimal documentation as an overview or general guidelines. This summary can help newcomers quickly grasp the software's purpose and give an outline of its structure.

Language and tools require significant investment in training licensing and support. The “best” language for the job is the one your current infrastructure supports and the maintainers who follow can understand

Software implementation is the phase where the software project begins to take shape, and it encompasses various aspects, including coding, testing, collaboration, quality assurance, security, and more. A successful implementation phase is crucial for delivering a reliable and functional software product that meets user expectations and project requirements.

Verifying code for aerospace engineering applications is often far more involved than in other industries. Code failure could mean destroyed vehicles and serious injury/death. Developing and documenting a testing suite and workflow should involve all technical stakeholders on projects, not just software engineers, so that diverse voices and can inform and review test results. If you have a hardware testing setup available (i.e., a FlatSat, a demonstration vehicle, hardware simulation tools, etc), testing both in virtual environments and on hardware is key to ensuring success, as one environment may reveal issues that the other wouldn't.

Software verification in aerospace engineering demands meticulous attention to detail and a rigorous approach to ensure the safety and reliability of critical systems. The verification process is a key element in achieving airworthiness and meeting the stringent safety standards required in the aerospace industry. Aerospace software must adhere to strict regulatory standards, such as DO-178C (Software Considerations in Airborne Systems and Equipment Certification) for civil aviation Along with other vanila checks

I recommend Test Driven Development (TDD) at this stage. I have explored TDD and experienced its many benefits to the whole software development cycle. TDD promotes a more disciplined and structured approach to development, resulting in higher-quality software that is more robust, maintainable, and adaptable to changing requirements. TDD brings together the developers, testers, product owners, and other stakeholders at the project's beginning to develop automated tests based on the requirements and the expected results. It follows the "Red-Green-Refactor" cycle to develop and verify the software iteratively. It encourages modularity, helps early detection of issues, and instills higher customer confidence.

Starting to apply the standard RTCA DO-178C-Software Considerations in Airborne Systems and Equipment Certification is absolutely a good choice.

Aerospace systems operate in various environmental conditions, such as extreme temperatures, pressure, and radiation. Validate the software's ability to function reliably in these environments through environmental testing and qualification.

While models are a great way to simulate your research, validation with empirical data sets is extremely important. Using real-world laboratory or field tests, your software will become robust and validated.

It's always a good idea to understand future maintenance resources to make the right design choices and avoid the pitfalls of unsupportable systems during the production phase.

Determine how your code will be maintained and used in the future. will it only be used by yourself? A team of researchers? Will it be open sourced? All of these influence the proper way to maintain your software.

While CI/CD can seem like something better suited for Silicon Valley setups or newfangled apps, well-defined CI/CD processes are key to well-documented and well-understood software testing. Automated testing and deployment can enable batteries of tests, of both hardware and software, to be run regularly and with minimal room for human interference (and mistakes). CI/CD pipelines running on each commit can let developers catch mistakes before spending hours hunting for bugs. All major Git hosting platforms offer integrated CI/CD that can often interface with test hardware - use it.

I agree with Michael B.'s comment on the importance of CI/CD. I think designing and developing a solid infrastructure pipeline using IaC tooling like Terraform at the beginning of the project is also essential, so infrastructure resources can be properly planned and evaluated, and cost properly assessed. Too often, infrastructure design/development comes at the last stage of the cycle, resulting in DevOps scrambling to throw something together to meet the project deadline. The design and development of the infrastructure pipeline and application CI/CD pipeline should be incorporated early in the project lifecycle, so DevOps doesn't become the bottleneck later in the project cycle.

Aerospace Software is very hard to test in physical environment - prototypes are very costly and cost of bug fixing is very high. Therefore for control software I recommend using Model Based Design approach in order to shift software testing as much as possible to the left in the development, at the best starting from requirement phase. Code, developed using MBD is also self-explanatory and easy to document.