X-Mule is an online game in which browser controls move a physical rover around a Mars-themed set. Rather than using a simulated world, actions such as charging and moving cargo take place with real robots and physical objects.

Players control the hardware over the internet and carry out missions inside a physically built arena.

Maker Mirza Makes modified a television remote so its number and control buttons can be used as a computer keyboard. The remote is an input device rather than the computer itself, and it was demonstrated with a Raspberry Pi 5-based handheld PC.

Inside the remote, the membrane keypad is connected to an MCP23017 I/O expander, which sends the button states to the Raspberry Pi over I2C. Rather than appearing as a conventional USB keyboard, the remote's buttons are read directly and turned into computer input.

Living directly under the departure path from San Francisco International Airport, Cameron Paczek built a Raspberry Pi 5 system that receives 1090 MHz ADS-B broadcasts with an RTL-SDR, calculates each aircraft's position relative to his home, and projects the live position onto his ceiling.

An outdoor PTZ camera uses ADS-B data for its initial aim, then YOLOX-Nano finds the aircraft in frame and adjusts the camera to follow it. The video is streamed to the TV, completing a plane-spotting setup that can be enjoyed from the couch.

Ollie Q hacked a Roomba so people anywhere can drive it in real time through a web interface. A low-latency video feed lets remote players steer around the room and compete to see who can clean most effectively.

A Raspberry Pi and custom software keep the controls responsive, and Ollie even experimented with a racing-wheel controller. A machine designed to automate housework is deliberately turned back into a human-controlled activity—just with the humans moved onto the internet.

Put transparent tape in front of a bare image sensor and the raw result normally looks like useless blur. Maker okooptics instead measures how that diffuser scrambles light and uses computation to reconstruct an image, then extends the same lensless approach into three-dimensional imaging.

Using a Raspberry Pi camera with a thin diffuser, different reconstructions of the same captured data can reveal views that behave like slightly shifted perspectives. Rather than adding more sophisticated optics, the project replaces the lens with something tape-like and transfers much of the difficulty into mathematics.

Press a button and the Vomit Comic Robot generates a one-off random comic and spits it from its mouth on receipt paper. Created by Cadin Batrack, the machine combines a Raspberry Pi, a thermal printer and a laser-cut plywood body. Processing software arranges hand-drawn elements into comic panels at random. There is no intended narrative relationship between the images, but the comic layout encourages people to invent one anyway. Even the physical output mechanism reinforces the joke: the robot appears to vomit nonsense comics on demand.

Bentobot was built so online friends could control real combat robots from phones and computers anywhere in the world. Thomas Nguyen created three small Raspberry Pi Zero 2 W-powered machines with drive motors and separate weapon motors, including flippers and spinning weapons. Viewers could join live streams, receive access through an app and immediately fight physical robots in the arena. Nguyen even created a lightweight custom language to reduce control latency. Instead of simulating robot combat online, the network ends at actual machines hitting each other.

Camera with Digital Film lets you physically pull the captured image out of the camera as a cartridge. A Raspberry Pi 3 A+ and 5 MP camera write each shot to a removable module containing a 1.54-inch four-color e-paper display.

The cartridge can then be displayed as a physical photograph. Because it is e-paper, the image remains visible without power and can later be overwritten. Modern digital hardware brings back one very specific analog pleasure: taking the picture out of the camera.

Whenever birds call outside, Fugleramme uses AI to identify them and displays only recently heard species as 19th-century illustrations on an e-paper frame. The Norway-based project runs BirdNET-Go on a Raspberry Pi 5.

Modern audio recognition works behind the scenes, while the visible result looks like an old natural-history book. Digital detections are transformed into a quiet, daily-changing bird guide for the immediate surroundings.