15— System Integration

00— Reflection

Below I have created a summary of some of the core aspects of how I am approaching system integration for the project Basin. These aspects and others that need to be retrieved from my subconscious have been guiding how the project takes shape. That said, I don’t believe it is possible to integrate a project after the first full development cycle. What are and are not critical design constraints often become apparent only after the full assembly is complete. Even if you try to design slowly and with a lot of care you simply can’t account for emergent details that only come into focus when the full work is brought into physical reality. Similarly, there are so many technical choices that can only be optimized after their full applied context is understood. In my case, my technical problem child turned out to be the PCB form factor. This large rectangle dictated a shocking number of design consideration that rippled across the entire design. Quite literally this single aspect was likely responsible for adding an additional 500-800 grams of material to the project. So to summarize, a project is never integrated until you design the whole thing again. Obviously, nothing is ever perfect, but the amount of things that can be improved by a full redesign can make it worth going through this process at least once… even though it is a huge pain in the ass.

02— The full System

Project Overview

Below (left) is a rendering of the full design. The rendering is unfinished as the holder for the dish has not yet been implemented nor have the couplers for joining the legs have been designed. There is simply too much work in CAD to have this finalized at this moment. Assembly and finishing are not the limiting factor for this design. The CAD modeling and 3D print time are primarily what limit the pace at which this project is developed. However, the core design and technical considerations I list below will be considered as these new components are developed.

Visual design constraints

  • The dropper assembly is designed as a circle to play off the circular motif found throughout the design.
  • Threes or multiples of threes are chosen wherever possible as there are three participants that take part in the work. This is why 3 instead of 4 legs were chosen. Similarly, 6 joining segments instead of 4 for the main body components.
  • The outer dropper assembly diameter is 400mm for design continuity with the dish below.
  • Flat walls are given a “fuzzy skin” texture that loosely resembles a roughly sewn cloth like burlap. This surface texture is applied through print settings. This texture breaks up some of visual monotony, softens the design, and connects it the lower fabric elements.
  • The project is not supposed to blend in, but represent a distinct space without walls. A circular rug is used to clearly demarcate the project from its surroundings.
  • Chrome, muted white, pastel blue, and black are used throughout the design with muted white and chrome being dominant colors and blue and black accent colors.
  • The form of the dropper funnels which blends sweeping curves with hard edges is another visual motif that is (once the missing components are designed) used throughout the design. This was to me the best compromise I could make visually to make the design feel cohesive.

Technical design constraints

  • The average human body needed to be considered throughout the design. This directly informed the height of the tent structure, sensor position, distances between participants, dish height etc.
  • The reservoirs (blue) were designed to easily hold minimum capacity of 1 liter.
  • A lot of care has been taken to discretely route wires throughout the project and insure that they can be accidentally pulled without creating undue stress on connection points. For most connections this solved by using off the shelf grove and JST-XH connectors.
  • I wanted a robust PCB design so I could decrease the chances of needing to redesign the board if aspects of the control system needed to change. To make the design more robust I based it around a very capable Xiao C3 and added a lot of additional IO. Similarly, Instead of designing two separate boards as was required by the distances between sensors and actuators I chose to create two identical boards to decrease cost and assembly time.
  • The rigidity of the tent structure has guided many decisions about mates between joining segments. Several options are available to me to solve this issue if the design proves to lack rigidity. Weighting the legs or concealing a plate under the rug are both viable options to address this potential problem.

03— Packaging

I am not concerned about packaging per se as the design is very large and is not intended to be a product. Packaging in the context of this work is more related to the end products that this project will produce in the form of media. So what this means is it is the photos, videos, posts and writing that represent how this physical work is delivered to people. That said, if I need to ship this project which is a possibility I will actually have to develop actual packaging system in the form of a custom crate. Again this is non-trivial in both design time and cost so this would happen at the very end of the design process if at all.

04—Schedule

01— My 5ish Week Plan

May 14→18

Focus: Mood boarding Initial CAD

May 19→25

Focus: Determine all supply needs and get working draft of dropper assembly assembled

May 26→1

Focus: Full assembly of armature and general polishing

June 2→8

Focus: Everything should be constructed care should be taken to photograph and document

June 8→10

Focus: You should be finished this should be used as a buffer period

June 12

Focus: Presentation