Author: Milosch Meriac
This week in our publicly streamed seminar, we presented our workflow for rapidly creating our hardware prototypes. We showed our process from schematics to pre-assembled Printed Circuit Boards (PCBs) and mechanical prototype production with iterative integration within less than two weeks.
We are confident that our workflow doesn't only apply to seasoned engineers but also for small junior hardware design teams to gain hands-on experience. Here are a few takeaways from our presentation:
In software engineering, Parkinson's law if Triviality is well known:
"The time spent on any [meeting] item of the agenda will be in inverse proportion to the sum [of money] involved."
An elegant solution to this social issue is Linus Torvald's approach - "Talk is cheap. Show me the code." – that is, forcing discussion participants to walk the talk and implement their claims instead of talking about it.
The "Show me the code" approach weeds out invalid solutions quickly and forces all discussion participants to deeply immerse themselves into the solution by working on it. Further, this approach removes discussion participants that are slowing down the solution process by not being ready to invest themselves into the solution. Work invested breeds respect by colleagues. It enables them to see other people's solutions they might have discarded themselves as infeasible.
Building multiple solutions for the same problem makes them comparable and simplifies any discussion tremendously. It does not mean to build them in production quality - just implementing them far enough to prove its ability for a carefully defined test case.
To transpose the success of "Show me the code" to hardware prototyping, we use these three techniques:
- We utilize quick turnaround PCB assembly services like JLCPCB to quickly try out electronic hardware ideas. Such services enable us to receive a fully assembled PCB within eight calendar days. Using tools like EasyEDA, we can design and submit PCBs for production and assembly between a few hours from having the idea to only a few days for complex projects. Like the prototyping step in software, the first PCB revision might only work for particular input signals and be incomplete - just enough for verifying ideas.
- While waiting for the assembled PCBs to return, we create dummy 3D models for "virtually" integrating them with our mechanical design ideas.
- Once we receive our PCBs back, we create functional mechanical solutions to enable testing under realistic (field-) conditions. The created mechanical designs can be limited in their functionality, durability, or required size constraints. Such an ad-hoc design process allows quick iteration over the possible solution space, quickly discarding ideas and narrowing down to the ideal solution.
- Findings from the iterative prototyping process are fed back into the design - resulting in more maturity of the first product releases.
Laser cutters are the ideal tools for trying out ideas quickly. Contrary to 3D printers, Laser cutters enable the big changes and re-spin of large mechanical designs painlessly within less than an hour.
Thanks to the low skill level required to operate laser cutters, remote collaboration and remote colleagues become efficient. They can produce your design idea themselves after minimal training, test them and iterate on them through modification.
Assembly and cloning of devices are further simplified by cutting out fixtures helping with the assembly process - further removing manual skill requirements for duplication.
In our Talk, we discuss approaches for increasing the precision of the assembled device various methods that apply to laser cutting:
- Helper devices like threaded six-sided aluminum cubes increase assembly precision.
- Design options like flexure hinges from the class of Compliant devices increase the achievable accuracy and simplify assembly, simplify sourcing, and reduce the overall part count.
While a seasoned engineer is critical in creating the production version of reliable and safe devices, it is beneficial to involve people from adjacent fields in the ideation fields. Typically, it would be a major nightmare to have people discussing topics outside their domain. By forcing everybody to work through the mechanical design in their own time or by prototyping with their colleagues, they immerse themselves with the engineering problems around it. Both can finally have a qualified and efficient discussion.
For researchers, basic skills and autonomy for creating simple assemblies and fixtures for their research allows them to falsify or verify novel ideas jointly with their colleagues - without tiresome sourcing processes and waiting weeks for specialized parts to arrive. After understanding and prototyping mechanical problems deeply, designs can still be sent to professional mechanical shops for production at much higher quality, precision, and maturity.
We recommend 2D and 3D design software with quick learning curves for non-engineers in our Talk, enabling researchers to share complex ideas with engineers. By communicating design ideas through 2D and 3D models, they can tear down the communication wall between them.
As a result, everybody involved will get a much better idea of the design constraints. Input from people in mechanical-engineer-adjacent fields is critical to success. Researchers and engineers can efficiently utilize feedback from user-interface design, graphical design, marketing, and software development at an early point using design-based communication. Thinking outside the box combined with deep experience in adjacent (but relevant fields) unlocks novel solutions and results in much better cross-field collaborations.