Intel Curie-Based Motion-Activated Lighting System

About this project

Goals

• Automatically turn on (and off) lights when footsteps are detected

• Synchronize the flashing of lights with the detected tempo of nearby dancing / tapping

• Output accelerometer/gyroscope readings over USB/Bluetooth for the detection of seismic events

GitHub Repo

https://github.com/ckuzma/arduino-101-sketches

Sketches

TapFlashTest

#include "CurieIMU.h" 
void setup() {
 pinMode(LED_BUILTIN, OUTPUT);
 CurieIMU.begin(); 
 CurieIMU.attachInterrupt(eventCallback);
 CurieIMU.setDetectionThreshold(CURIE_IMU_SHOCK, 1050); // 1.050g = 1050mg 
 CurieIMU.setDetectionDuration(CURIE_IMU_SHOCK, 75);   // 75ms 
 CurieIMU.interrupts(CURIE_IMU_SHOCK); 
} 
void loop() { 
 // We don't need to put anything in the main loop... 
} 
static void eventCallback(void) { 
 if (CurieIMU.getInterruptStatus(CURIE_IMU_SHOCK)) { 
   digitalWrite(LED_BUILTIN, HIGH); 
   delay(50); 
   digitalWrite(LED_BUILTIN, LOW); 
 } 
} 

• Flashes onboard LED when a tap is detected

• Used to figure out changes to impulse detection

TapBeatMatch

• Synchronizes flashing of onboard LED to the tempo of tapping on tabletop surface near the Arduino 101

• Implemented weighted average to help mitigate noise

• Adjustable shock detection threshold

SeismicReader + Python Visualization Script

• Arduino 101 board outputs JSON-formatted accelerometer and gyroscope values over Serial, which is then read by a connected computer

• Wrote a small Python script that reads data from the board and graphs it on very simple lines in the terminal / command prompt

• Saves results to a CSV that can be opened in Excel and graphed

• Multiplatform Python 2.x / 3.x usage:

VibrationLighting

• Originally designed to detect approaching footsteps and activate lighting while someone is nearby, was scaled back to double-tap detection as the signal-to-noise ratio wasn't clean enough for my installation

• LED strip taped underneath the entryway mirror turns on and off when the mirror is knocked on twice in a row

Bonus

A conversation with Bogdan made me realize that it'd be very easy to modify the Arduino sketch to work with standalone accelerometer + gyroscope sensors connected to nearly any Arduino device, and then use CurieGraph.py to visualize the data. Seeing as the goal of collecting seismic readings is to prove that it's possible to massively crowd-source data without complicated or specialized equipment, I rummaged in my desk for the "ITG-MTU" / "GY-521" / "MPU-6050" sensor board, and wrote the following Sketch for it. CurieGraph.py might fail to start a couple of times, but usually it catches on properly by the third attempt.

ExternalGyroscope

#include<Wire.h> 
const int MPU_addr=0x68;  // I2C address for ITG-MTU 
int ax, ay, az, temp, gx, gy, gz; 
void setup(){ 
 Serial.begin(9600); 
 Wire.begin(); 
 Wire.beginTransmission(MPU_addr); 
 Wire.write(0x6B); 
 Wire.write(0); 
 Wire.endTransmission(true); 
} 
String jsonEncodeValue(String key, float keyVal){ 
 return "\"" + key + "\":" + String(keyVal) + ""; 
} 
String assembleJson(String keysAndVals){ 
 return "{" + keysAndVals + "}"; 
} 
void loop(){ 
 Wire.beginTransmission(MPU_addr); 
 Wire.write(0x3B); 
 Wire.endTransmission(false); 
 Wire.requestFrom(MPU_addr, 14, true); 
 ax = Wire.read()<<8|Wire.read();   
 ay = Wire.read()<<8|Wire.read(); 
 az = Wire.read()<<8|Wire.read(); 
 temp = Wire.read()<<8|Wire.read(); 
 gx = Wire.read()<<8|Wire.read(); 
 gy = Wire.read()<<8|Wire.read(); 
 gz = Wire.read()<<8|Wire.read(); 
 // temp = temp/340.00+36.53; // Convert temp data to celsius - NOT BEING USED 
 String keyVals = jsonEncodeValue("ax", ax) + ","; 
 keyVals += jsonEncodeValue("ay", ay) + ","; 
 keyVals += jsonEncodeValue("az", az) + ","; 
 keyVals += jsonEncodeValue("gx", gx) + ","; 
 keyVals += jsonEncodeValue("gy", gy) + ","; 
 keyVals += jsonEncodeValue("gz", gz); 
 Serial.println(assembleJson(keyVals)); 
 delay(100); 
} 

Code

arduino-101-sketches