P1 · IoT

Indoor air-quality monitor

Sensors breathe room air, not board heat.

White air-quality monitor with an e-paper display standing on a wooden bookshelf in morning light

Process: FDM → Injection
Quantity: 10 → 1,000
Region: Global
Environment: Indoor
Image: Concept image, not to scale

Example brief · not client workWe wrote this brief ourselves; it is not client work. The design below was generated and checked by our engine as an internal test; read the notes under it before drawing conclusions.

The request, as a client would send it

Hi, we're a small team building an indoor air-quality monitor for home offices and classrooms. The electronics work on the bench and we need a proper enclosure: 10 printed units for a pilot, then about 1,000 injection-molded if the pilot goes well. What goes inside: - ESP32-DevKitC V4, 48.26 x 27.94 mm (the antenna sticks out about 6 mm past the board) - Sensirion SCD41 CO2 sensor on our own 25 x 25 mm carrier board - PMS5003 particle sensor, 50 x 38 x 21 mm, fan inlet and outlet on one short side - Waveshare 1.54" e-paper module, 48 x 33 mm - USB-C power only, no battery The main problem with our breadboard version: the CO2 sensor reads high whenever the ESP32 is busy, because it sits right next to it. The sensors have to breathe room air, not board heat. For size and feel we like how compact the Aranet4 is. Ours will be bigger because of the PM sensor, and it must not look like theirs. AirGradient's open-source ONE is a good example of the airflow idea. Our rough sketch is attached. It should work on a desk and on a wall. Matte white, and no visible screws on the front would be nice. Please send STEP files, drawings and a BOM; we'd like to order prototypes within three weeks.

Attached by the client

Concept sketch of the air monitor with its lid lifted, showing separate sensor and board compartments — Concept sketch · stands in for the client's own sketch · AI generated

Links in the request

ESP32-DevKitC V4 dimension drawing (opens in a new tab)48.26 x 27.94 mm PCB
Sensirion SCD41 (opens in a new tab)10.1 x 10.1 x 6.5 mm sensor
PMS5003 (Adafruit 3686) (opens in a new tab)50 x 38 x 21 mm
Waveshare 1.54" e-paper module (opens in a new tab)48 x 33 mm outline
Aranet4 HOME datasheet (size reference) (opens in a new tab)71 x 71 x 24 mm; do not copy the look
AirGradient ONE (open-source airflow reference) (opens in a new tab)130 x 130 x 35 mm

Product names are their owners' trademarks and are used only to describe size or parts.

Highlighted numbers came with the request; they are the values we would design to.

What we considered

No.	Consideration	How it is checked
01	Sensor air kept away from board heat An internal wall splits the sensor chamber from the ESP32; room air enters low and leaves high, so the CO2 and PM sensors never sit in the board's warm plume.	Wall present; both sensors placed inside the sensor chamber; ESP32 outside it.
02	The particle sensor's fan can breathe The PMS5003 inlet and outlet face open grilles with a clear gap, not the housing wall.	Clear zone in front of the sensor's inlet and outlet.
03	A real USB-C plug reaches the port The opening is sized for a plug's overmold (about 12.5 x 7 mm), not just the receptacle.	Clear path in front of the receptacle for the plug body.
04	One shape for printed pilot and molded run Uniform walls and screw bosses that work printed now and molded later; draft angles are recorded as assumptions for the tooling stage.	Wall thickness range; boss sizes; draft listed on the brief card.

The question we would ask you

Is it mostly on a desk or mostly on a wall?

It decides which face carries the display angle, where the vents sit, and whether the stand or the keyhole is the primary mount.

What we do not claim

Measurement accuracy of the sensors
Airflow or thermal simulation results

From request to checked design

1 · Request
Sensors breathe room air, not board heat.

Plain-language brief with 6 links and a sketch. Only the text goes into the pipeline.
2 · Brief card
2 parts to make4 parts to fit16 pass criteria

Assumed
- Quantity 1000 moulded per posting; 10 FDM prototypes first.
- ABS matte white chosen — stable, easy to texture; ASA if sunlight exposure matters.
- Wall 2.2 mm nominal uniform (posting states none) — typical for ABS enclosures this size.
- and 4 more
Asked
- Should the e-paper be flush behind a clear window or recessed in an open bezel?
- Is wall mounting via keyhole acceptable, or do you need a separate mounting plate?
- and 1 more
3 · Design
Injection moulding (10 printed first)

front shell × 1 · back shell × 1

Engine output. Untextured CAD render.
4 · Check

16checks passed

0checks failing

9checks not measured

Measured
- Closed solids
- Overall size
- Part fits with clearance
- Part inside the envelope
- Clear zone in front of a part
- Cavity size
- Holes and bosses
- Wall thickness
- No two parts overlap
- Part can be removed
- Printing: overhangs
Counts cover checks that exist; the notes below list what has no check yet.

Failures go back to the design stage with the numbers; earlier failures stay on record.
5 · Scene
AI-generated from the CAD render above: same shape, only material, light and background added. Not to scale.

Read before you trust it

The screw bosses were enlarged after a person approved the change.
The fixer changed the screws from M2.5 to M2 without confirmation; drawings and BOM must follow that or revert it.
The draft-angle check found no faces to measure, so it is counted as not measured; the final bosses have no draft.
Eight moulding and printing checks are not built yet and are counted as not measured.

Run on 2026-09-29 with our in-house engine as an internal test. Drawings and BOM for this case are not produced yet.