Electron Dust built the Hexa-Bouncer to keep a ping-pong ball bouncing on a moving platform. Two cameras track the ball while six arms move and tilt a triangular platform to meet its predicted landing point.
According to the creator, the machine uses a Teensy 4.0, six geared NEMA 23 stepper motors, six motor drivers, a 36 V 14 A power supply, two cameras and a GPU-equipped Windows PC. The image-recognition and motor-control software are custom-built, and the many aluminum parts were machined on a desktop CNC.
Keeping one tiny ping-pong ball going calls in two cameras, six powerful motors and a GPU. The support equipment completely dwarfs the thing it is supporting.
Mark Rober invited engineers and makers to CrunchLabs Casino to build and test devices for cheating at casino games. Cheating itself was allowed, but getting caught was not.
The creations included a remote-controlled robotic poker chip that could move a roulette bet after the wheel stopped, dice whose internal weighting could be changed with heat, and a concealed wearable card-counting assistant. Another contestant hid a pinhole camera in a sleeve and used computer vision to reconstruct the order of a deck.
Change the rule from “no cheating” to “cheating is fine if you don't get caught,” and a casino suddenly becomes an engineering contest.
SAM is a flying robot inspired by the autorotating motion of a falling maple seed. It uses a single motor to spin its whole body for lift, with the same actuator responsible for both staying aloft and controlling direction.
Researchers at the Singapore University of Technology and Design added predictive control and tested the robot on circular and figure-eight paths, under fan-generated wind, and with a 5 g payload. Rather than adding the multiple rotors used by conventional drones, the team improved accuracy while keeping the machine to a single moving part.
Instead of adding rotors like a normal drone, the whole robot just keeps spinning like a maple seed. The mechanics got simpler while the control problem got harder.
Steven Gong and fellow University of Waterloo engineering students built a robot that can retrieve a shirt, straighten it, fold it, and place it on a stack. Two human-operated leader arms control two follower arms while cameras record the process; the team collected about 200 demonstrations for training.
Early models either failed to begin the task or moved violently enough to break a 3D-printed gripper. The team eventually fine-tuned a robotics foundation model, then had to collect more data after moving the setup to a new building changed the lighting. By the next morning, the system could complete the folding sequence autonomously.
A chore that takes a person less than a minute required four robot arms, hundreds of demonstrations, a broken gripper, and an overnight model search. Laundry has excellent scaling properties.
Dynamic Creatures launched in September 2026 to turn advanced mobile robots into interactive characters for theme parks, hotels and other guest-facing spaces. The company is Boston Dynamics' official entertainment and hospitality partner and plans to build both character systems based on existing robotic platforms and fully custom machines.
Its target sits between fixed animatronics and task-oriented service robots. Public examples show familiar-looking quadruped robotics transformed with fur, faces and character design into something closer to a roaming theme-park creature. After solving difficult locomotion, one of the next engineering problems is apparently making the machine look huggable.
After years of improving robot-dog locomotion, the next technical requirement turns out to be “make it fluffier.” Robotics has successfully advanced all the way to costume design.
Keeping one tiny ping-pong ball going calls in two cameras, six powerful motors and a GPU. The support equipment completely dwarfs the thing it is supporting.