12 Sensor Project Ideas for Industrial Design Students (Portfolio Edition)
Twelve portfolio-ready sensor project ideas for industrial design students, from breath-responsive lamps to radar-activated mirrors, with parts lists and notes on what each one demonstrates.
Portfolio reviewers see a lot of lamps. They see a lot of chairs. They see a lot of side tables with a tastefully imperfect veneer and an essay about intentionality.
They do not see many objects that respond to a breath. They do not see many objects that know whether they're being held. They remember the ones that do and they remember the student's name.
Physical computing, for an ID student, isn't about becoming an engineer. It's about adding one more verb to your portfolio (responds, listens, senses, adapts) next to all the nouns everyone else is showing. The object doesn't have to be complicated. It has to communicate an idea you can't communicate with form alone.
Here are twelve projects that land well in an ID portfolio, roughly sorted from "doable in a weekend" to "worth a whole semester." Each one has a parts list you can actually buy and a short note on what it proves about you as a designer.
Don't just copy these. Look at them, pick the idea that rhymes with the brief you actually have and bend it into your own.
1. The breath-responsive lamp
A desk lamp that gets brighter when you exhale toward it.
What you need: INMP441 i2s microphone, a XIAO ESP32-S3, a WS2812B LED ring (24 pixels is plenty), a 3D-printed diffuser in whatever shape makes sense for your concept. Optionally, a small BME680 to distinguish breath from ambient air.
What it demonstrates: Subtle interaction. Not a button, not a touch, but a gesture you already make. Reviewers notice the thoughtfulness of the input.
The gotcha: Getting the microphone sensitivity right takes time. Calibrate to the ambient room noise, don't just trigger on peak loudness, or it'll react to anyone talking nearby.
2. The weight-aware desk organizer
A tray that knows which object you just picked up, because it's watching how the weight on it changes.
What you need: Four load cells in the corners (sold as a "50kg half-bridge" set) plus an HX711 amplifier per cell, or one load cell at the center plate for simpler projects. A Pi Pico works well for fast ADC reads.
What it demonstrates: Invisible input. No cameras, no microphones, just pressure. People find this surprisingly magical because they can't see how it works.
The portfolio angle: This is one of those projects where the video is the deliverable. Film from above, in good light, show objects being placed and removed and the organizer reacting. The video is the portfolio piece.
3. The ceramic touch surface
A glazed ceramic vessel (or tile, or bowl) with invisible capacitive touch zones underneath the glaze.
What you need: An MPR121 capacitive touch chip (it handles 12 channels), thin copper tape or painted conductive ink and a ceramic piece with a flat-ish bottom. Glue or epoxy the copper to the underside of the ceramic. The chip senses through the ceramic just fine, because glazes aren't conductive.
What it demonstrates: Material literacy. You're showing that you understand capacitive touch works through non-conductive materials, which lets you preserve the material honesty of the ceramic while adding interaction. This is an ID thesis waiting to happen.
Practical note: Do a small test tile first. Some glazes with metal oxides interfere. Better to find out on a tile than on your final piece.
4. The gesture-controlled pendant lamp
A pendant you turn on by waving a flat hand under it, dim by holding your hand still at a distance, turn off with a swipe.
What you need: An APDS-9960 gesture sensor. That's basically it, plus whatever drives the light. The APDS-9960 detects up, down, left, right and distance changes out of the box. The library is excellent.
What it demonstrates: A complete interaction vocabulary. Four gestures, four functions. Nothing on the object to look at, no hardware you can see. The lamp looks like a lamp.
5. The tilt-aware bedside object
A small object (a clock, a lamp, a sound player) that changes behavior depending on how it's oriented. Upright: alarm mode. On its side: quiet. Upside down: do-not-disturb.
What you need: An MPU-6050 or LSM6DS3 IMU, a small e-paper display (if you're doing the clock version), any microcontroller. The whole BOM is under $25.
What it demonstrates: Physical affordances as interface. You're turning an object's orientation into a control, which is something a graphical UI cannot do and a physical object can.
Why it works in a portfolio: It's a concept that photographs in one or two clean shots. Upright beside the bed. On its side. Done. Reviewers absorb it instantly.
6. The scent diffuser that reacts to the calendar
An aroma diffuser that changes its scent blend based on the time of day, the weather, or your calendar. Morning: citrus. Afternoon: something neutral. Evening: something warmer. Cancelled meeting in twenty minutes: something calming.
What you need: A small piezo-driven ultrasonic diffuser (you can buy the piezo disk bare and design the reservoir yourself, which is the fun part), an ESP32 for WiFi and either small solenoid valves or small pumps for scent switching. This is the most mechanically ambitious project on the list.
What it demonstrates: Ambient technology. Scent is the least-used sense in UX and ID and doing it well is rare.
Honest warning: Mechanical scent switching is hard. If you can only do one scent that varies in intensity instead of composition, that's still a good project.
7. The sound-reactive acoustic panel
A wall-mounted acoustic panel whose visible texture (pattern of LEDs behind fabric, or small moving elements) responds to the FFT of the sound in the room.
What you need: INMP441 i2s microphone, an ESP32 (the S3 has an FFT-friendly CPU), a grid of WS2812B LEDs or a string of them routed into a pattern, acoustic felt for the facing material.
What it demonstrates: The invisible made visible. This is a classic ambient-display brief and the acoustic panel framing gives it a functional reason to exist on a wall instead of being "art."
FFT tip: Use the arduinoFFT library for small signals. You don't need to understand the math to make it work. Just feed it your samples and read back the frequency bins. It's one of those libraries where the first run usually works.
8. The posture coach chair
A chair backrest that senses how you're sitting and gives quiet haptic feedback when you've slumped for too long.
What you need: An IMU embedded in the upper back of the seat (or two IMUs, one in the seat and one in the back, to measure the angle between them), a small vibration motor, a Pi Pico or XIAO, a rechargeable LiPo, a charge circuit.
What it demonstrates: Wellness product design with a working prototype. You can use this brief in a human factors studio, an ergonomics class, or a wellness brief.
The hard part: Figuring out what "good posture" actually means in terms of sensor readings. Don't hard-code one "correct" angle; let it learn from the first two minutes of the user sitting normally, then flag deviations from their baseline.
9. The footfall-counting rug
A rug, runner, or doormat that counts how many people have walked across it today and visualizes that count somewhere.
What you need: A grid of force-sensitive resistors or an array of piezo pickups sandwiched between two layers of felt or fabric. A multiplexer if you have more than 8 sensors (the CD74HC4067 16-channel mux is under $3). An ESP32 for the wireless display or the web dashboard.
What it demonstrates: Data from the physical environment. This is exactly the kind of project that pairs well with a brief about retail spaces, museums, schools, or housing.
Honesty note: FSRs are not very consistent. You get relative data, not absolute counts. Lean into that. This is for a sense of how busy a space is, not an exact count.
10. The plant that asks for water (the good version)
Yes, everyone has seen this project. The reason to include it is that almost no one does it well. The standard version uses a capacitive soil moisture sensor and an LED that turns red. Done, boring.
The good version is an object (a little figurine, a ceramic form, a printed character) that becomes its state. It droops when thirsty using a servo hidden under the pot. It chimes softly once an hour until you water it. It remembers when you last watered it. It has a personality.
What you need: Capacitive soil moisture sensor (the corrosion-resistant kind, not the two-prong one), a small servo, a XIAO ESP32-S3 so you can log data if you want, an enclosure you designed to move.
What it demonstrates: Character design through motion. You're treating the object as a character, not a widget. This is more ID than engineering and that's why it lands in a portfolio.
11. The door handle that knows
A door handle (printed or cast or machined) that knows whether the door is currently open, closed, or being pushed. It can light up from the inside, or send a signal to something else, or just record the state.
What you need: A small magnet in the door frame, a hall effect sensor in the handle, an accelerometer in the handle to detect pushing or pulling. Power it from a small coin cell and only wake up when the IMU detects motion (this is the trick that makes the battery last for months).
What it demonstrates: Sensors disappearing into the object. The handle looks like a handle. A reviewer who doesn't know to look for a chip will never know one is in there.
Portfolio shot: A cutaway render showing the internals, next to a photo of the closed object. The contrast is the story.
12. The mirror that wakes up when you approach
Not the Pepper's Ghost kind with a screen behind glass. The real one. A mirror that's just a mirror until you walk up to it, at which point a subtle light around the edge fades on. Walk away, it fades off.
What you need: A 60GHz mmWave radar module (LD2410 or similar, about $8 and honestly magical, or a simple PIR if you want to start easier), an ESP32 or a Pico, WS2812B LEDs around the mirror edge, a diffuser and a real mirror.
What it demonstrates: Radar is the underused sensor of the decade. It sees through thin walls, works in total darkness, doesn't care about light, doesn't feel creepy the way a camera does. Using it correctly is a signal that you've looked beyond the obvious sensor choices.
Why it's a great thesis project: Mirrors are loaded with meaning. A mirror that "notices" you opens up a lot of narrative directions. Calm, eerie, intimate, clinical. Same hardware, different contexts, different portfolios.
A few things worth saying about every one of these
Film it before the review. Every project on this list is an object that does something and a photo can't show doing. Get a friend to hold your phone while you interact with the object, in good natural light and film it once working all the way through. This video is the portfolio piece. The object is a prop you carry so nobody thinks you faked the video.
Design the object first, then wire it. Not the other way around. If the brief is a ceramic vessel, the ceramic comes first and you figure out how to fit the electronics inside. If the electronics determine the shape, it's going to look like an engineering project with a skin on top and reviewers will notice.
Pick one idea and make it excellent. Do not try to build three of these for one deliverable. Build one, film it, document it, explain why it exists. The portfolio page for one done-right interactive object will beat three half-finished ones every time.
Save your sensor data. Even if you don't use it, log it. Later, when you're writing your portfolio page, you'll want to talk about what you observed: how many times the rug was stepped on, how the breath sensor responded to different distances. Data makes the page feel real.
Your parts list is a design decision. In the portfolio writeup, don't hide the electronics. Don't pretend they're not there. Show the BOM, show the enclosure cutaway, show the wiring diagram next to the final photo. This is your work. Own all of it.
None of these projects require you to be a great programmer. A few hours with the example sketches for each sensor will get you most of the way. The hard part is the ID part: the choice of object, the choice of gesture, the choice of what the object means. That's the part you're already being trained to do.
The electronics are just how you make the idea move.
If you hit a wall on any of these, the Arduino mistakes ID students make before a review post probably covers whatever went wrong. And if your tooling is slowing you down, the best tools for ID students building interactive prototypes piece is the shopping list somebody should have handed you in second year.