My Fab Academy capstone project
I am developing TrailNAV so that a hiker can save a location and see the direction and distance back without mobile data. The proposed system combines GPS, heading input, an embedded controller, a display and solar-assisted battery power.
The build currently documents the earlier printed enclosure, component fitting and wire preparation. I have also prepared a new 1.3-inch OLED design and first navigation firmware. Navigation, final custom electronics and power performance still need a complete operating test. This page records that progress and the intended next steps.
The original large display is not available, so I am moving to a 1.3-inch, 128 × 64 I²C OLED. I am using SH1106 as the starting driver choice, with an SSD1306 option in the firmware. I still need to identify the exact module and check its pin order, voltage, outside dimensions and mounting holes.
I could not recover the editable model of the earlier enclosure. I prepared a new parametric Fusion design with Codex assistance instead. The native file contains the sketches, dimensions and feature timeline, and the generator script is included so I can rebuild the exports. The starting enclosure is 82 × 128 × 33 mm, including a 3 mm faceplate.
I can rotate the assembly, remove the cover in the interior view, or separate the parts in the exploded view. These views come from the actual Fusion exports. They show the mechanical parts, without pretending that the electronics have already been installed.

The printed base has a 2.4 mm floor and walls, corner screw bosses, and four supports for a removable electronics tray. The tray has cable-tie slots so I can adjust the module positions after measuring them, and a clearance notch behind the save button. A separate printed backing frame supports the OLED behind the faceplate. The laser-cut faceplate provides a subtractive part for this revision; it has not been cut yet.
I reserved space below the tray for a battery envelope of 50 × 30 × 7 mm. This is a layout allowance, not a confirmed battery size or a finished battery restraint. The charger, GPS, compass, connector access and final PCB still need to be matched to the actual parts. I am keeping the power test seperate from the enclosure fit.

I prepared a first software version for the XIAO RP2040. It reads GPS sentences, checks that the fix is fresh, saves one waypoint in flash and shows distance and the bearing back. A short press saves the first point. Once a point is stored, I need to hold the button for two seconds and release it to replace that point.
The displayed bearing is relative to true north. This version does not read the compass or show which way the enclosure is facing. If a fix is lost, it hides the live navigation values and retains the saved waypoint. The RP2040 build and software calculation checks pass, but this does not establish an outdoor test or prove retention through a real power cycle.
For my first bench check I will use USB power, identify the display and GPS connections, and confirm that the pins are accessible on my board. Battery and solar operation come after the charging and regulation circuit is settled. I still need to make the final project PCB and test the complete device.
I will print the small screw-fit coupon and compare the actual OLED with the faceplate drawing before making the full enclosure. The screenshots and interactive viewer record the CAD work; fabrication, charging and navigation results will be added after the physical checks.
I designed enclosure parts with openings for the display and solar panel, prepared them in the slicer and printed them. The CAD views and photographs below show that process. The original editable enclosure model and saved fabrication settings still need to be archived for a repeat build.
The design tools are introduced in Computer-Aided Design, and the fitting sequence is documented in System Integration.
The intended signal path is GPS and compass → controller → waypoint storage and display. A control would save the current location, and the firmware would retrieve that position for the return-navigation view. The battery, charger and regulation need to match the load and the panel.
The midterm diagram shows the original functional proposal. The planning list names a XIAO ESP32-C6 and E-Ink display, while the construction material and screen-test sketch show different hardware. I need to resolve those identities before publishing an as-built pin map.
| Planned component | Qty | Recorded unit price (RWF) | Recorded line price (RWF) |
|---|---|---|---|
| XIAO ESP32-C6 | 1 | 20,000 | 23,600 |
| u-blox NEO-M8N GPS | 1 | 19,682 | 19,682 |
| QMC5883 compass | 1 | 4,500 | 5,310 |
| 2.7-inch E-Ink display | 1 | 27,080 | 27,080 |
| 3.7 V battery | 1 | 9,000 | 10,620 |
| Solar panel, described as 12 V / 100 mA | 1 | 8,900 | 10,502 |
This is the provisional planning budget, using the prices in my original notes rather than current quotations. The recorded line prices sum to 96,794 RWF. They do not include the charger/regulator, custom PCB, connectors or enclosure material. The differences between some unit and line prices still need an explanation, so the subtotal is not a verified final build cost.
The sourcing references and part discrepancies are described in Week 17.
I designed and fabricated an RP2040 board during Electronics Design and Electronics Production. Those pages link the board files and its input/output test. The final TrailNAV controller-board revision has not yet been established.
The archived TFT screen-test sketch draws patterns and a TrailNAV splash screen. Its GPS and heading labels are static text; the sketch does not read location or heading. The new OLED source adds GPS input, fresh-fix handling, waypoint storage and display updates. Compass-heading processing and a physical navigation demonstration remain to be completed.
My final electronics record will connect the actual schematic and PCB layout to the fitted modules, power circuit, board settings and firmware dependencies. That will make the build reproducible rather than relying on an external view of the assembled device.
I fitted the screen and solar panel to the printed enclosure and prepared wires for the electronics. The photograph below records the panel fit. The dedicated controller and battery mounts, connector arrangement and secured wire routes still need to be documented.
The next assembly needs clearance for the cover, support for the modules and access for servicing. I will test each subsystem before and after mounting so that changes caused by the enclosure can be identified. The new OLED design includes a laser-cut faceplate as its proposed separate subtractive part. It still needs to be cut and fitted; the PCB milling record covers electronics production.
I will check each function before combining them in an outdoor demonstration. The following tests define what the prototype needs to show. No measured navigation, charging or runtime results are assigned to the device yet.
| Function | Planned acceptance check |
|---|---|
| Location | Acquire a valid outdoor fix, record coordinates and fix information, and compare with a known reference. |
| Waypoints | Save a location, restart the device and retrieve the same waypoint. Show distance and direction changing during movement. |
| Heading | Calibrate the compass and compare readings at known directions, both outside and inside the enclosure. |
| Offline use | Complete the save-and-return workflow without a network connection. |
| Power | Measure load and charging behaviour under recorded conditions using the actual charging circuit. |
| Runtime | Run a timed battery test with a recorded display update rate and operating mode. |
| Packaging | Operate with the cover closed, secured internal parts and accessible controls. |
The final demonstration needs to use the assembled device and its intended power source. It will show the saved waypoint, its retrieval and changing navigation information during movement, alongside the power and packaging results.
| Stage | Work and dependency |
|---|---|
| 1 — Confirm the hardware | Identify the actual controller, display, GPS, compass, battery and charging circuit. Resolve their voltage and interface requirements before finalizing connections. |
| 2 — Build and check the electronics | Complete the board design and fabrication, then test its power rails and individual input/output connections. |
| 3 — Implement the navigation loop | Acquire valid location and heading, store a waypoint and drive the actual display before closing the enclosure. |
| 4 — Complete the physical assembly | Finish the project’s subtractive part, mounts, connectors and wire routing after the component interfaces are settled. |
| 5 — Test the assembled device | Run the waypoint, offline-operation, charging and runtime checks with the intended power source and enclosure. |
| 6 — Publish the completed build | Add the final sources and results, update the presentation and review the project with the instructors. |
The wider proposal, prior-work comparison and budget are in Week 17. The dissemination plan describes how I intend to share the working build and gather feedback.
The summary slide and construction video introduce the idea and the enclosure work. The video does not yet show a complete navigation workflow. I will update it with the operating demonstration after the functional tests.
Download the presentation material and the archived screen-test source below. The original image and portrait construction video are retained alongside the landscape versions.
My prior-work comparison identifies navigation, solar-charging and GPS-parsing references. The final firmware and circuit record will also identify any sources actually reused and the changes I made.
The shared machine work is linked from Week 12. Albert UWURUKUNDO assisted with the plasma postprocessor described in Week 16. Claude helped generate the starting website, and Codex helped edit the written documentation and prepare downloadable examples; the website-writing scope is described in Week 1. For the OLED revision, Codex also generated the Fusion API design script, navigation starting code and browser viewer; I still need to check these against the actual hardware.
I selected MIT for my original code and CC BY 4.0 for my original written documentation, as explained in Week 18. Design-file terms and third-party attribution remain separate. Assessment guidance is available in the Fab Academy final-project requirements.