Week 13 — Mid Term Review

Week 13 was not a new fabrication week for me. I used it as a checkpoint: read back through the earlier assignments, mark weak documentation, and turn the final project from a rough idea into a build plan I could actually follow after midterm.

What I reviewed

I checked the first half of the site against the assignment pages, not only against my memory of what happened in the lab. The main pattern was clear: the projects worked, but some pages needed stronger evidence, clearer process notes, or better source-file links.

Week	Status at midterm	What I needed to fix or keep tracking
Week 1	Website and project introduction started.	Keep the final project description aligned with the newer Brainfog direction.
Week 2	CAD work documented with early enclosure models.	Make sure later 3D printed shell photos link back to the original design files.
Week 3	Parametric laser-cut kit and vinyl cutter workflow documented.	Keep kerf settings and construction-kit result visible, since they are easy to miss in a long page.
Week 4	Embedded programming page had board setup, serial control, and button test.	Make the serial protocol explicit so it reads as programming protocol work, not only a blink test.
Week 5	3D print and scan were documented.	Add a clearer note on 3D printing limits, especially supports, overhangs, and fit.
Week 6	KiCad schematic, layout, DRC, and source archive were present.	Keep the Week 8 production page linked so the designed board is not isolated from the milled result.
Week 7	CNC chair design, CAM, machining, and G-code were documented.	No major gap, but the dogbone and slot-clearance decisions should stay easy to find.
Week 8	Electronics production page showed milling, stuffing, and first firmware test.	Make the workflow checklist explicit: toolpath, milling, stuffing, debugging, programming.
Week 9	Input device page showed encoder wiring, sensor logic, and live readout.	Keep the AMS1117 workaround clear because it explains why the wiring is different from a clean board.
Week 10	Output device page showed LCD wiring, I2C notes, contrast adjustment, and test video.	No major gap. The page already ties the display to the board I made.
Week 12	Machine build page had the vending-machine story and my integration role.	Add clearer technical problem solving and next-design improvements.

Final project proposal at midterm

My final project is a desktop brain-fog support device for people who feel tired, unfocused, or mentally slow during study or work. The system combines a small desktop host, a camera path for eye-open and eye-closed state, a wearable sensor path, voice interaction, screen feedback, and a simple intervention flow. I am not building a medical device. The goal is a Fab Academy prototype that notices fatigue-related signals and gives gentle support.

Part	Midterm direction	Why it belongs in the project
Desktop host	3D printed shell on a laser-cut platform.	It holds the screen, speaker, button, and main electronics in one place.
Eye-state sensing	XIAO ESP32S3 Sense with a local model for open and closed eye state.	Blink count gives one simple signal related to attention and fatigue.
Wearable path	Small wrist module sending body-signal data back to the host.	The device should not rely only on the camera.
Voice and screen interface	Button-triggered conversation, screen status, and audio feedback.	The interaction needs to work without a laptop beside it.
Intervention	Short breathing, rest, or 40 Hz audio support.	The output should be a small action the user can try immediately.

Project progress

At midterm, I had separate pieces, not a finished product. That was useful to admit. The next phase needed fewer isolated tests and more integration work.

Area	Already done	Still open after midterm
Mechanical design	Early host shell models and printed enclosure tests.	Final platform layout, camera tower, wearable charging position, and bottom closure.
Electronics	Small milled XIAO carrier board from electronics design and production weeks.	Main project PCB, WROOM wiring, power path, speaker, microphone, storage, and camera link.
Input and output	Encoder and LCD tests proved that my board could read input and drive output.	Move from weekly test modules to final project input and output devices.
Firmware	Serial command structure and page-based LCD test firmware.	Combine sensing, screen, audio, storage, voice, and intervention logic.
Documentation	Most weekly pages had photos, videos, and source files.	Tighten missing evaluation points and keep source files downloadable.

Project management plan

I changed the plan after this review. Instead of trying to make every feature perfect at once, I split the rest of the semester into build passes. Each pass has one thing to prove.

Pass	Work	Exit condition
1. Core electronics	Confirm WROOM host pin map, microphone, screen, speaker, button, storage, C3 bridge, and XIAO S3 camera link.	All modules run on the bench with known pins and shared ground.
2. Mechanical enclosure	Finish laser-cut platform, 3D host shell, camera tower, switch position, and rear power access.	Electronics can fit inside without blocking screen, speaker, camera, or charging port.
3. Sensing loop	Bring eye-state data and wearable data into the host firmware.	The host can show current state without needing the serial monitor.
4. Interaction loop	Add voice trigger, screen response, speaker output, and simple intervention flow.	One user session can run from button press to feedback.
5. Integration and documentation	Move from breadboard to PCB, close the product form, record final tests, and update weekly pages.	The final project page links to source files, wiring, photos, and demo video.

Review result

This review made the project less vague. Before midterm I had a direction and several tested skills. After midterm I had a sequence: finish the host electronics, lock the enclosure, bring in sensing, then build the interaction loop. The documentation plan also became clearer. If a page does not show the process, the test, and the source file, I need to fix it before final review.