Streamlining Hardware Development: The Role of Design of Experiments (DOE) and a Plan of Record (POR)

‍

Design of Experiments (DOE) is a systematic approach to experimentation that helps engineers optimize and fine-tune the performance of complex systems. Essentially, a DOE is a structured framework for conducting experiments and gathering data by varying multiple factors simultaneously. The primary objective of a DOE is to identify the optimal set of parameters that result in the desired product performance.

A DOE involves several key steps:

1. Problem Definition: Engineers define the problem they need to solve and the performance metrics they want to optimize.
2. Variable Selection: They select the factors (variables) that may influence the performance. Often times these variables are tolerance changes, mechanical updates, electrical changes, or swapped alternate OTS components that impact the system.
3. Experiment Design: Engineers design a set of experiments that systematically vary these factors. The goal of getting these experiments designed is to not vary a system too much such that the impact cannot be traced back to a change. As we’ve learned in elementary school, it’s best to vary one variable at a time to directly correlate end impact.
4. Data Collection: Conduct the experiments, collect data, and analyze the results. Often times early data collection efforts may lead to learnings that drive even more DOEs. The data collection process should be consistent across DOEs to ensure accurate analysis. Generally DOEs take place during pre-defined prototype builds, but can also occur ad-hoc
5. Optimization: Identify the optimal combination of factors that yield the desired outcome.

‍

‍

Plan of Record (POR): The Blueprint for Hardware Development

Once engineers have optimized their hardware using DOE, they establish a Plan of Record (POR). POR is essentially the blueprint for the final hardware design that captures all the optimized parameters and configurations. It serves as the reference for all subsequent stages of development, including manufacturing and testing.

Key elements of POR include:

1. Design Specifications: Clear and precise documentation of the hardware's design, including schematics, drawings, and technical specifications.
2. Testing Procedures: Instructions for testing and quality control procedures to ensure the final product meets performance requirements.
3. Bill of Materials (BOM): A detailed list of all components and materials required for the system.
4. Timeline and Milestones: A timeline for the various stages of development and milestones to track progress.

Managing Multiple DOEs: A Complex Challenge

In the world of hardware development, engineers often face the challenge of managing multiple DOEs concurrently. This arises because hardware systems are complex, with numerous interdependent components and variables. Each DOE may focus on different aspects of the system, such as power consumption, speed, or reliability.

Current Practices in Managing DOEs

Traditionally, engineers have managed multiple DOEs using a combination of spreadsheets, documents, and project management tools. While these methods have been effective to some extent, they come with several limitations:

1. Version Control: Tracking changes to the DOE and POR documents can be cumbersome, leading to confusion and potential errors.
2. Documentation Overhead: Maintaining detailed documentation for each DOE can be time-consuming and may divert engineering resources from core tasks.
3. Collaboration: Collaboration among multiple team members can be challenging, especially when working on different aspects of the system.
4. Visibility: It can be difficult to gain a holistic view of all ongoing DOEs and their progress.

The Paradigm Shift: Branching and Merging in Hardware Development

Here is where the parallel with software engineering becomes evident. Software developers have long embraced the concept of branching and merging to manage complex codebases efficiently. In branching and merging, each feature or bug fix is developed in its own branch, and then these branches are merged back into the main codebase when they are ready.

Applying this methodology to hardware development can offer several benefits:

1. Isolation of Changes: Each DOE can be treated as a separate branch, isolating changes and configurations from each other. This reduces the risk of unintended consequences driven by the parent-child relations within a complex system.
2. Parallel Development: Engineers can work on different DOEs concurrently, just as software developers work on different features in separate branches. Each branch can maintain it’s own documentation and record keeping.
3. Version Control: Like software version control systems (e.g., Git), hardware development teams can use version control tools to track changes and manage different versions of each DOE. This increases traceability and makes design changes easier to understand.
4. Merge Points: Establishing defined merge points in the hardware development process ensures that optimized parameters and configurations are appropriately integrated into the POR.
5. Traceability: By associating each DOE with a specific branch, engineers can easily trace back changes, understand their impact, and manage dependencies.
6. Collaboration: Collaboration becomes more streamlined, as engineers can work independently on their designated branches and merge their changes into the POR as needed.

In the ever-evolving field of hardware development, achieving optimal performance while managing multiple DOEs is a complex challenge. Engineers rely on DOEs to optimize their systems and establish a POR as the blueprint for the final design. However, the traditional methods of managing multiple DOEs come with limitations.

The adoption of branching and merging workflows from software engineering can offer hardware development teams a powerful solution to these challenges. By treating each DOE as a separate branch and merging them into the POR when ready, engineers can streamline collaboration, enhance version control, and improve traceability. This paradigm shift not only brings greater efficiency but also empowers hardware developers to meet the demands of increasingly complex projects in a rapidly advancing technological landscape.

Ready to evolve your hardware development process to modern workflows? See how Bild’s branch and merge feature can help!

‍