compareeHub

THE PIXAR LAMP IS REAL - 260 DOLLARS IN PARTS AND IT LOOKS AT YOU WHEN YOU TALK

A desk lamp that turns to look at you when you talk, stretches when it wakes, and runs entirely on open-source code — inspired by the Pixar intro and Apple's ELEGNT paper.

by Binh Pham / Human Computer Lab

FULL CAD BOM FIRMWARE DOCS

RoboticsHome

difficulty
●●●○○
time
a long weekend
license
GPL-3.0
repo
repo ACTIVE540 stars

WHAT YOU’LL NEED

  • 3D printer + filament — printable parts — files are in the repo
  • Electronic parts — full list with part numbers in the repo BOM
  • Dev board / microcontroller — runs the project firmware
  • Soldering kit + basic tools — one-time buy, serves every future build

Partner

Reach your Pi from anywhere — Remote desktop to the machine running this build — no port forwarding, works right from the browser.Try Getscreen.me
Videos
1

COMPAREE VERDICT

LeLamp is one of the most characterful open robotics projects to come out this year, and the build is more approachable than the feature list suggests. It runs on a Raspberry Pi 4 and six serial-bus servos (the SO-101 open-hardware actuator base plus one extra for the head tilt), with a USB camera and microphone for vision and voice. The emotion system uses an LED ring and animated poses — stretching at wake, turning toward whoever spoke, drooping when idle — all driven by a straightforward Python stack that uses OpenCV for face tracking and a voice activity detector to know when to look up. The one thing most likely to go wrong is sourcing the servos: the bill of materials links to Taobao and some AliExpress sellers that may not ship globally, and substituting servos means rewriting the serial protocol code unless you stick to the exact ST3215 and ST3020 models specified. If you can get the servos and are comfortable with basic Python and a bit of Linux setup (installing dependencies, running scripts at boot), this is a weekend-plus build that will make everyone who sees it smile. If you have never wired a servo bus or run code on a Pi, start with something simpler — the docs assume you know what /dev/ttyUSB0 means.

NOT IN THE REPO

  • —Full 3D-printable STL files for all structural parts (11 pieces total)
  • —Complete Python codebase for vision, voice, emotion system and servo control
  • —Detailed bill of materials with Taobao and AliExpress links (some may be China-only sellers)
  • —Assembly guide with photos and wiring diagrams in the README
  • —GPL-3.0 licence — allows commercial use with source disclosure
  • —No PCB design included; servos and electronics are off-the-shelf modules wired together

Can I build this?

Print11 STL files: base, arm segments (three pieces), two joints, lampshade, head mount, cable guides. Roughly 18-24 hours of print time on a standard FDM printer.
BuyRaspberry Pi 4 (2GB or better), six serial-bus servos (five ST3215, one ST3020), USB camera module, USB microphone, 12V power supply (3A minimum), WS2812B LED ring (16 LEDs), DuPont wires, M3 screws and heat-set inserts. Total approximately $260 if buying from China; expect $300-350 if sourcing locally.
ToolsFDM 3D printer, soldering iron (for heat-set inserts and any wire joins), screwdriver set, basic Linux command-line skills.
SkillsIntermediate: comfortable with Raspberry Pi setup, Python package installation, and serial communication. The assembly is straightforward once you understand how the servo bus works, but debugging a non-responding servo or a misconfigured camera means reading logs and checking connections.
Time8-12 hours of active work: 2-3 hours printing (you can batch overnight), 2 hours assembly, 3-4 hours software setup and testing, 1-2 hours tuning poses and tracking behaviour. Spread across a long weekend.
Cost$260-350. The servos (six at ~$15-20 each) and the Raspberry Pi 4 (~$55-75) account for most of it. The camera, microphone and LED ring are under $30 combined if you buy generic modules.

Build at your own risk. Projects involve tools, electronics and sometimes mains voltage — follow the creator’s safety notes.

A desktop 3D printer that prints in full colorPartner · Kickstarter
A desktop 3D printer that prints in full color

HeyGears G1: 10M+ colors and transparent parts in one print, plus UV printing on flat objects. Figures, parts and labels — no painting.

See how it prints

Videos

LEGO Ideas Disney Pixar Lamp Set REVEAL

No official video. The repository includes a short demo clip embedded in the README showing the lamp tracking a person's face and responding to voice.

Gallery

https://opengraph.githubassets.com/12edff538afc9136503fdb9e26d2652d75ab34573538781570fbcaa55c6f78f9/humancomputerlab/LeLamp
https://img.youtube.com/vi/CUvns7igEQM/maxresdefault.jpg
https://img.youtube.com/vi/fWGD3zz6YlE/maxresdefault.jpg
https://img.youtube.com/vi/S4NBWqFS3io/maxresdefault.jpg

Start here

Navigation into the creator’s own docs — we don’t rewrite the guide, we route you to the source.

  1. 1.Read the full README on GitHub (The README contains the complete bill of materials, wiring diagrams and software setup instructions. Start there to confirm you can source the servos.)
  2. 2.Print all 11 STL files(Standard PLA is fine. Install heat-set inserts while parts are warm off the bed.)
  3. 3.Order the servos and electronics (The BOM includes Taobao and AliExpress links. If those sellers do not ship to your region, search for ST3215 and ST3020 serial-bus servos elsewhere — but verify the protocol matches before buying substitutes.)
  4. 4.Assemble the mechanical structure (Follow the photo guide. The servo horn alignment matters — mark the zero position before tightening screws.)
  5. 5.Install the Python environment and dependencies (Uses OpenCV, pyserial, sounddevice and a few other libraries. The README provides a requirements.txt file.)
  6. 6.Test each subsystem individually(Run the servo test script to confirm all six servos respond, then test the camera feed, then the microphone. Fix one thing at a time before running the full emotion loop.)

Resources

Documentation, files and community threads for this build — we link straight to the original sources and never rehost the creator’s files.

Blu-ray drive, 10 Gbps hub and card reader in one boxPartner · Kickstarter
Blu-ray drive, 10 Gbps hub and card reader in one box

BluHub reads and burns Blu-ray/DVD/CD, adds USB-C, USB-A, SD/TF and 10 Gbps data — plug and play, grand-piano finish.

See the hub

KNOWN ISSUES

  • The servos are the single biggest sourcing trap: the BOM links to Chinese marketplaces and some sellers will not ship internationally. If you buy different servos, you will need to rewrite the serial protocol code unless they happen to use the exact same Feetech STS command set.
  • The Raspberry Pi 4 needs adequate cooling. The vision and voice processing are not intensive by modern standards, but running them continuously without a heatsink or fan will throttle the CPU and make tracking laggy.
  • The servo bus uses a single shared data line (half-duplex serial). If you have never wired one before, the fact that TX and RX are bridged can be confusing — read the servo datasheet and the wiring diagram carefully before powering on.
  • Pose tuning takes longer than you expect. The default angles in the code are a starting point, but every printer's tolerances and every assembly's friction are slightly different — budget an hour to tweak the stretch, nod and idle positions until they look natural.
  • The camera must have a clear view of the user. Mounting it inside the lampshade sounds clever but kills the field of view — follow the reference placement in the photos.
  • The voice activity detection is basic (threshold-based RMS). It works in a quiet room but will false-trigger on background noise. If your workspace is loud, you may need to adjust the threshold or add a better VAD library.

Can I use different servos if I cannot source the ST3215 and ST3020 models?

Yes, but only if the replacement servos support the same Feetech STS serial protocol. If you switch to a different protocol (e.g. Dynamixel, plain PWM), you will have to rewrite the servo control layer in the Python code. The repository does not include alternative drivers.

Does the camera or microphone capture anything permanently?

No. The vision system processes frames in memory to detect faces and the voice system listens for activity level only (no speech recognition, no recording). Nothing is saved to disk unless you modify the code to do so.

Can I run this on a Raspberry Pi 3 or a Pi Zero?

The README specifies a Pi 4 with at least 2GB of RAM. A Pi 3 will struggle with the OpenCV face detection at a usable frame rate, and a Pi Zero will not run it at all. Stick to the Pi 4 or upgrade to a Pi 5 if you already have one.

What is the power consumption?

The six servos under load (e.g. all moving at once during the wake-up stretch) can draw close to 3A at 12V. The 12V-to-5V converter for the Pi and the LED ring add another 1-1.5A peak. Use a 12V 5A supply if you want headroom; a 3A supply will work but may sag under full load.

Is there a way to add speech synthesis so the lamp can talk back?

Not in the current release, but the architecture would support it — you would add a TTS engine (e.g. pyttsx3 or an API call to an external service) and trigger it from the emotion state machine. The hardware has no speaker, so you would need to add a USB or I2S audio output module.

Community builds

No community builds yet — be the first, we feature the best ones.

Discussion1

FROM THE COMPAREE TEAM

Six servos, a Pi 4, and about 18 hours of print time — and you have a lamp that stretches when it wakes up. If you have built one, did you stick with the ST3215 servos or swap them for something easier to source?

CompareeTEAM27d agoedited

Practical notes from our verification: the repository is actively maintained (last push July 2026), the README is detailed and the bill of materials is complete with direct links to parts, but almost all the servo and electronics links point to Taobao or Chinese AliExpress sellers — international builders will need to hunt for the same models from local suppliers or be prepared to wait for slow shipping. The single biggest practical upgrade is not in the hardware, it is in the pose tuning: the default angles are a starting point, but every build will need an hour of tweaking to make the motions look natural rather than robotic. The repository includes a short demo video showing face tracking and voice response, but there is no official walkthrough or assembly timelapse — you are working from the README and photos alone. That said, the documentation is clear and the author responds to issues, so if you get stuck on the servo protocol or camera setup, you can ask.

Binh Pham / Human Computer Lab

Binh Pham built LeLamp as part of research at the Human Computer Lab at the University of Chicago, inspired by Apple's ELEGNT paper on expressive robotics and the classic Pixar lamp character. The project demonstrates how much personality you can encode in a simple five-DOF mechanism when you combine vision, voice and a well-tuned animation system.

GitHub

Star the project on GitHub

DISCLAIMER

  • Comparee is not the author of the projects featured here. All rights to each project belong to its creator — every page links to the original source, and we never host creators’ files.
  • Information is provided without warranty and may become outdated as projects evolve. Prices are indicative bands only — always check the creator’s parts list for current costs.
  • Building and operating any project is at your own responsibility. Protective equipment, safe workshop practice and compliance with local regulations are the builder’s responsibility.
GitHub repo