Published November 21, 2017

The Intrudalert

The Intrudalert creates an invisible IoT turnstile, which can be used for theft prevention and store monitoring.

IntermediateShowcase (no instructions)4 hours648

Things used in this project

Hardware components

Particle Photon

Solderless Breadboard Half Size

PIR Motion Sensor (generic)

Jumper wires (generic)

USB-A to Micro-USB Cable

Story

This project was created for the UNCC MEGR 3171 IOT Project.

If you have ever seen the movie Bad Santa, you know that store security can be easily tricked by intruders if they are able to hide in the store once it closes. Our team has come up with a "invisible turnstile" that can be implemented in stores or businesses to keep track of the amount of people that are inside.

Part 1

Utilizing IoT and the expansive capabilities that it holds, our team designed a measuring system that is able to count the number of people that enter and exit a store, with the ability to observe and analyze this data in real time with internet connected devices.

The main task of the Intrudalert is to alert store employees if there is still someone in the store at closing time. At the end of the business day, the amount of people in the store should be the exact amount of employees that are still present. If you count more people than the number of employees, then that means that either someone left through the roof, or, the more likely scenario, there is still someone in the store. Once this information is known, the employees can then check the typical places where customers might be, such as the dressing room or bathroom. But, if these locations are inspected, and no one is found, then an intruder is likely present, and preventative actions can be taken to prevent theft.

The secondary task of the Intrudalert is to capture real time data of the store's activity. Need help deciding how many employees you need and when they should work? Use the Intrudalert to determine the peak times in the store when it is the busiest. Want to improve customer satisfaction? The Intrudalert notifies when a customer enters the store, and an available salesperson can immediately greet the customer at the entrance.

Part 2

This is the Intrudalert set up in our house, it was collecting data on the number of people in our living room.

This is the Intrudalert set up between the kitchen and the living room.

Another angle of the Intrudalert set up, showing the high traffic area that was being measured.

Each of the two PIR sensors were positioned to trigger when someone walks into the field of vision, which is defined by the angle of the binder opening. It is important that these fields of view don't overlap which could lead to an incorrect In/Out signal.

Over the course of an hour data was gathered regarding the number of people that were in our living room. The gathered data was sent to Thingspeak and is graphed in a useful format that would be perfect for any small store to monitor the number of people that were in there store throughout the day.

The data captured during an hour period in the living room.

Video

This video shows us talking through the set up and demonstrating how this can be used.

How to set up the Intrudalert along with a demonstration.

Code

bool alertedin;
bool alertedout;
void alertin();
void alertout();

// PIR motion detection is running off interrupt functions, which once triggered
// will run either "alertout" or "alertin" functions. Pins D1 and D2 were
// selected as signal inputs from the PIR sensors.
void setup() {
    pinMode(D1, INPUT);
    pinMode(D2, INPUT);
    attachInterrupt(D1, alertout, RISING);
    attachInterrupt(D2, alertin, RISING);
    alertedin = false;
    alertedout = false;
}

// publishes an event to Particle Console when either the outside or
// inside PIR sensor is triggered
void loop() {
    if (alertedout == true) {
        Particle.publish("MotionOut",String(Time.now()),21600,PUBLIC);
            alertedout = false;
    }
    if (alertedin == true) {
          Particle.publish("MotionIn",String(Time.now()),21600,PUBLIC);
            alertedin = false;
    }
}

// sets booleans to true to tigger publish event in running loop
void alertout() {
    alertedout = true;
}

void alertin() {
    alertedin = true;
}

int outtime;
int intime;
int diff;
int i;
bool out;
bool in;

// subscribes the receiver Photon to the published events from the
// mesauring Photon. In the last arguement for the subscribe function, enter
// the device ID from the publishing Photon.
void setup() {

    Particle.subscribe("MotionOut", a,"ID from Publishing Photon");
    Particle.subscribe("MotionIn", b,"ID from Publishing Photon");
    out = false;
    in = false;
    i = 0;
}
// converts Unix time data from subscribe functions to an integer value
void a(const char *event, const char *data) {
    outtime = atoi(data);
      out = true;
}

void b(const char *event, const char *data) {
    intime = atoi(data);
      in = true;
}

// calculates whether the set of triggers was a person entering or exiting.
// Publishes activity to Particle Console
void loop(){
  if (out && in) {
    diff = outtime - intime;
      if (diff > 0) {
        i--;
      } else {
        i++;
      }
    Particle.publish("Count",String(i),21600,PUBLIC);
    out = false;
    in = false;
  }
}

Credits

Cameron Palkowski

1 project • 1 follower

Robert Cook

1 project • 1 follower

The Intrudalert