This week answers four linked questions about the final project — the shoulder-clip posture corrector: what's actually inventive here, how the open work gets out into the world (dissemination), who owns and can build on it (intellectual property), and how the commercial version sustains itself (income / business). I treat all four as design choices — the same way I treated the PCB track width or the BLE / ESP-NOW radio split — there are trade-offs, and I want them explicit.

Use of AI:
this srea was new on me and i needed alot of gidance and information, but I also needed to finishh theis as fast as possible and get back to finlizing the technicalities with the final project. I use Claude AI and this was the main prompt I used.

01: Two-track model: open for education, reserved for commerce

---

The single most important decision on this page is that the project lives on two tracks at the same time:

This hybrid is the same open-core / commercial-premium pattern that Arduino, Adafruit, and Pimoroni use. Nothing in Fab Academy's open-documentation ethos prevents a student from later building a commercial spin-off using new IP — the academy work is what made the spin-off thinkable in the first place.

02: Dissemination plan (open track)

---

For the open track, the default position is open hardware and open source: the schematic, PCB layout, firmware, and app blocks are public on my Fab Academy repo and on this site. The reasons:

• The project depends entirely on prior open work — KiCad, MicroPython, MIT App Inventor, ESP-NOW examples, the FA archive, and dozens of community tutorials. Closing the result would be inconsistent with how it was built.
• Posture biofeedback is well-explored commercially (Upright, Lumo Lift). My niche — dual-shoulder symmetry inspired by yoga and horse-riding posture — is small enough that the educational value is in adaptation, not in protecting the schematics.
• An open release lets future Fab Academy cohorts pick this up — academy projects are most useful as training material when others can fork and modify them.

How the open track ships: the GitLab repo (link from W01), the final-project page as the official write-up, the process page for the build narrative, and a short demo video from W16 showing the working system. Stretch channels: an OHWR submission of the hardware design files, a short write-up for the FabLab Network blog, and posting the App Inventor .aia to MIT's gallery.

03: Invention and prior art

---

What's actually inventive here, and what isn't? I split the project component-by-component against the standard "useful + novel + non-obvious" lens:

Component	Inventive?	Prior art / why
ESP32-C3 + MPU6050 sensor stack	No	Commodity building blocks — used in hundreds of FA archive projects and thousands of maker tutorials. Standard MicroPython driver.
BLE-gateway + ESP-NOW relay (left clip = bridge to right clip)	Implementation novelty — not patentable	ESP-NOW is Espressif's own protocol; the gateway pattern is documented in Espressif application notes. My contribution is choosing it over two parallel BLE sessions for this specific dual-shoulder use case — a design decision, not a discovery.
Dual-shoulder symmetry framing for posture correction	Conceptual contribution	The product idea — correcting posture by tracking shoulder symmetry rather than a single spine point — is the project's distinctive angle. Builds on prior commercial work (Upright Go = single sensor, Lumo Lift = single clavicle sensor) and on Fab Academy archive work (Nadine Uwineza = flex sensors for spine; Praveen Kumar = generic ESP32 + MPU6050 integration). The framing is novel, but it's not patentable subject matter — it's a design choice grounded in my yoga and horse-riding background.
Pause-and-place magnet pocket in 3D-printed enclosure	No	Well-known assembly technique in the maker community — documented in Bambu and Prusa community guides.
The integration story end-to-end	Editorial contribution	The W01 → W17 documentation chain — how a vest concept simplified through 17 weeks of spirals into a working two-clip BLE+ESP-NOW posture corrector — is the project's main reusable contribution. Openly licensed as CC BY-SA so the next cohort can use it as curriculum.

Prior-art / freedom-to-operate notes: a quick search on Google Patents for "posture correcting wearable vibration motor" returns a wide commercial field — mostly full-back or single-spine devices, none specifically targeting dual-shoulder symmetry. The implementation reuses very common open building blocks, so I'm not pursuing patent protection. This is a maker-level scan, not a rigorous freedom-to-operate analysis — a proper one would be needed before launching the commercial product.

04: Intellectual property and licence stack

---

The work splits into three open categories and one reserved category, each needing its own licence:

Asset	Track	Licence	Why
Documentation, photos, schematics, drawings on this website	Open	CC BY-SA 4.0	Anyone can share, remix, and build on the writing — with attribution and share-alike, so derivatives stay open.
Firmware (MicroPython on the ESP32-C3) + MIT App Inventor blocks	Open	MIT	Permissive — low barrier to entry for other students and makers. No copyleft burden downstream.
Hardware design files (KiCad schematic, PCB, Gerbers, 3D-printable STL, MDF DXF)	Open	CERN-OHL-P v2 (Permissive)	Designed specifically for open hardware — disambiguates how hardware-design files (vs. code or text) are licensed; recognised by the open-hardware community.
Commercial product (productised PCB, injection-moulded housing, multi-language native app, wireless-charging case, branding)	Reserved	All Rights Reserved © Hamidah Rahimi 2026	Developed as a separate IP track from the academy artefacts. The open licences above explicitly do not extend to the commercial product. Trademarks for any future product name are reserved.

Operational checklist — what makes the licence choice real:

• Add a LICENSE file at the repo root pointing to CC BY-SA 4.0 (catch-all default).
• Add a LICENSE-MIT file in the firmware/ directory.
• Add a LICENSE-CERN-OHL-P file in the hardware/ directory.
• Add an SPDX licence header (// SPDX-License-Identifier: MIT) to every source file.
• Add an ATTRIBUTIONS.md listing every reused open source — KiCad, MicroPython, the MIT App Inventor BLE extension, ESP-NOW example code, and any tutorial whose code I borrowed.
• Add a footer line on every page of this site: "Open work © 2026 Hamidah Rahimi — CC BY-SA 4.0. Commercial product reserved."

05: Commercial product roadmap

---

The commercial track turns the academy artefact into a productised wearable.

• Target customers: people who need to correct their posture — anyone with long office hours, on-site construction or building work, drivers, students, yoga and mindfulness practitioners.
• Value proposition: dual-shoulder symmetry feedback, plug-and-play, multi-language. Different from Upright (single sensor, spine) and Lumo (single clavicle clip) because the symmetry signal is something only two coupled sensors can produce.
• Price point: $100 USD for the complete kit — 2 clips + carry case + multi-language app additionanl subscribtions model for advanced app features.
• Cost path to $100: the academy W17 BOM comes to ≈ $160 at one-off / Amazon retail. Going commercial, the key parts (ESP32-C3, MPU6050, vibration motor, LiPo, TP4056) become 60–80 % cheaper at order quantities ≥ 100 from LCSC / Alibaba; switching from FDM print to injection-moulded clip housing (per-unit) is cheaper at scale; replacing the laser-cut MDF carry case with a moulded shell collapses two manufacturing steps into one. The combined effect brings the unit cost below $40 — making $100 retail viable with comfortable margin.
• Productisation steps over the academy version:
  • SMD-package MPU6050 directly on the PCB (fixes the drop-test failure documented in W15).
  • Injection-moulded clip housing replacing FDM PLA.
  • Multi-language app (English + Arabic minimum; more by demand).
  • Wireless-charging carry case (the W17 §05 Spiral Development feature).
  • Native Android / iOS app replacing the App Inventor build.
• First customers: the FabLab community — both as early buyers (they already understand the value of a thoughtfully-designed wearable) and as multipliers (instructors who can recommend it to their students).
• Channels: direct sales at first; later — B2B for compeies employees- yoga studios, physiotherapy clinics, makerspaces; an open-kit listing alongside the commercial product (the two coexist — the open kit is for builders, the commercial product is for end users).

06: Business Model Canvas

---

The commercial-track business model in one place — using the classic Strategyzer Business Model Canvas (Osterwalder, 2010). Nine blocks arranged so the operating side (Key Partners → Key Activities → Key Resources) sits on the left, the customer-facing side (Customer Relationships → Channels → Customer Segments) sits on the right, the Value Proposition bridges them in the middle, and Cost Structure + Revenue Streams sum the choices made above. This video explains the relation between the different blocks

07: Presentation deliverables

---

The week's deliverables are a slide and a short video.

• Summary slide — single image: hero shot of the two clips + app on a phone, project name, one-line problem, one-line solution, the two-track model (Open: CC BY-SA / MIT / CERN-OHL-P • Commercial: All Rights Reserved at $100), and contact.
• Video — ≤ 1 minute. Calibration → wear → slouch → vibration alert → reset, with the app screen mirrored in a corner. Compressed with the FFmpeg workflow from Week 02.
• Pitch script — 30-second elevator-pitch version of the project for the live presentation, in plain English. WIP