Punch-Activated Flame Thrower Glove

Context & Motivation

This was a personal project, and at its core, it came from a simple motivation: I wanted to make something undeniably cool. Something that catches the eye. Something that stops someone in their tracks and makes them ask, "Is that what I think it is?"

I wanted to build something visceral. Something that feels powerful and immediate. The absence of buttons was intentional—it makes the interaction feel instinctive rather than mechanical.

My Role

I independently designed and built the glove as a fully integrated wearable system. I handled mechanical design, embedded electronics, firmware development, gesture recognition logic, and complete system integration from concept to working prototype.

How It Works

An onboard IMU (Inertial Measurement Unit) continuously measures acceleration and orientation in real-time. I developed custom processing logic that distinguishes an intentional punch from everyday arm movement using a combination of directionality analysis, timing windows, and acceleration thresholds.

When a valid punch gesture is detected, the system triggers a tightly synchronized sequence of actuation. The fuel delivery system activates, the ignition fires, and flame bursts from the glove in immediate response to the motion. All components are housed inside a custom 3D-printed glove structure designed for compactness, durability, and wearability.

Motion Detection Algorithm

The core challenge was distinguishing intentional punches from normal hand movements. The algorithm analyzes multiple parameters simultaneously:

• Acceleration Magnitude - Threshold-based detection of rapid forward motion
• Directional Vector - Confirms motion is primarily forward (punch trajectory)
• Angular Velocity - Detects wrist rotation characteristic of punching motion
• Timing Window - Validates acceleration profile matches expected punch duration
• False Positive Filtering - Rejects ambiguous gestures to prevent accidental activation

Real-Time Processing

The microcontroller samples IMU data at high frequency (100+ Hz) and processes it through the gesture recognition pipeline. When confidence exceeds the threshold, the system executes the actuation sequence with sub-100ms latency, creating the perception of instant response.

Mechanical Design & Fabrication

3D-Printed Structure

The glove housing was designed in CAD software and 3D-printed using heat-resistant materials. The design prioritizes compactness while maintaining accessibility for component mounting and maintenance. Living hinges and snap-fit joints allow for assembly without external fasteners.

Fuel System Integration

The fuel delivery system required careful routing to prevent line kinking while allowing full hand articulation. I used flexible tubing with strategically placed support points that move with the natural motion of the hand. The nozzle positioning ensures safe flame direction away from the wearer.

Wearability Considerations

Weight distribution was critical for comfortable wear and natural movement. Heavier components (battery, fuel reservoir) are positioned on the forearm, while lighter electronics sit on the back of the hand. Adjustable straps ensure secure fit during rapid motion.

Firmware Development

Sensor Fusion

The firmware combines accelerometer and gyroscope data to build a complete picture of hand motion. Complementary filtering reduces sensor noise while maintaining responsiveness to rapid movements.

State Machine Architecture

The system operates through a well-defined state machine: idle monitoring, gesture detection, actuation, cooldown, and return to idle. This structure ensures predictable behavior and prevents rapid re-triggering.

Calibration & Tuning

Extensive iteration was required to find the right balance between sensitivity and reliability. Threshold values were calibrated through repeated testing, adjusting parameters until the system reliably detected punches while rejecting false positives from normal movement.

Engineering Philosophy

From an engineering standpoint, the challenge was turning something flashy into something reliable. Gesture-based control is messy and ambiguous, and making it feel confident requires careful tuning and iteration. This project taught me how expressive systems demand more rigor, not less.

The absence of buttons wasn't just an aesthetic choice—it fundamentally changed the interaction model. When the interface is invisible, when your body is the controller, the system has to be absolutely predictable. There's no room for latency, no tolerance for false positives. The technology has to fade completely into the background, leaving only the pure relationship between gesture and response.

Project Reflection

This project represents the intersection of spectacle and precision engineering. What appears as pure showmanship—fire controlled by a punch—is built on careful sensor fusion, real-time processing, and obsessive attention to safety. The goal was to create something that feels magical while being completely deterministic under the hood.

Building systems that respond to human gesture taught me that the most compelling interfaces are often the most invisible. When technology disappears and all that remains is intention and response, you've created something that feels less like a tool and more like an extension of yourself. That's the real achievement here—not the flame, but the feeling of control.