How to prototype Connected Devices
Case study: iLock — Smart lock that opens with facial recognition
About
iLock is a smart door lock service that helps user keep their home safe, while facilitating with the ease of interacting people at the doorstep.
The lock comes with a smart door bell which works with facial recognition. It can also be operated remotely using the app. It has two-factor authentication to keep home more secure. The lock predicts behaviors of people entering home and give access recommendations.
Team: It’s a solo project.
NOTE: This is only a prototype that I created using Arduino, motor, sensors etc. So in the demonstration video, you may see the intestines of the door lock floating outside.
Research
Note: This is only a prototyping project, hence the research was quick and dirty. And may/will be skewed.
Secondary Research
From the quick secondary research, I explored the “security” domain.
Where all do we need our environment around us secure?
Sensors exploration:
IR distance sensor (10–80cms) for facial recognition prototyping, Capacitive touch sensor for locking/unlocking door, Ambient Light sensor for light intensity control, Motion sensor for detecting movement in the house, Earthquake sensor for alarming about the earthquakes
Interviews
Then, after interviewing 3 people about home security, I came to conclusion that:
- People forget keys, and then someone else has to come and open the door
- People feel the home is safe if one family member is always there at home.
- Mostly people keep their valuables at bank lockers
- Smart security devices themselves are not secure enough. They can easily be hacked and collected information out of them (address, whats inside home, names and phone numbers of family members, record videos of what’s happening inside home)
Insights and opportunities
The research suggested opportunities in home security. Later, after exploring smart locks like Lockitron, August lock, Goji, Unikey, Haven, August camera etc. I came to conclusion that many smart locks do exist, but they are either difficult to install, need improvement in usability, bulky in size, or hang up.
Concepts
Feedback control system
Form Exploration
In order to use facial recognition, I needed to build the lock that has some element that goes on the outside of the door. But I also wanted the lock installation super easy. Therefore I connected the camera and facial recognition sensor (IR sensor for prototype purpose) in the door bell and connected that with the lock. Bell is created out of black foam.
Foam Board
I used black foam board to create the front section of the lock and base of the door bell. White foam was used to create the door bell button.
Styrene Sheet
I used this material to create the handle of the lock as the material was hard enough to hold tightly and soft enough to turn it into a cylinder.
A Jar Cap 😐
I also used the cap of a jar that goes inside the lock. This cap supports the handle to remain in shape. Also, inside the lock, it mounts the motor on one side and bulb on the other
Technical Exploration
Arduino Uno
I used Uno to connect the whole circuit. Ardiuno code runs on the logic of distance of person from IR sensor (Door bell) and based on that turns on the bulb and rotates the motor and eventually opens the lock.
Sparkfun Wifly Shield
The Arduino is mounted by Wifi shield that helps it connect with the smart phone via Blink. I downloaded Wifly libraries and inserted its code in my arduino code and linked it to Blink app.
IR distance sensor
In order to showcase the feature of facial recognition, I used an infrared distance sensor and connected it to 5V on arduino. Hence I faked it for prototype sake. IR sensor goes into the Door bell so that whenever person rings the bell or comes close, it opens the lock (if the person has the access) or notify’s the owner.
Motor
I connected the motor to 5V through bred board to arduino. Motor runs on the logic of distance of a person from the door bell.
Links: Framer + Audio , Arduino + HomeKit integration, Arduino + Wifly shield, notifying phone from arduino using blink,
Arduino Code (without Wifly & Blink installation)
#include <Servo.h>int sensorpin = 0; // analog pin used to connect the sharp sensor
int ledpin = 13; // analog pin used to connect the sharp sensor
int val = 0;
int newval = 0;
// variable to store the values from sensor(initially zero)
Servo myservo;
void setup()
{
Serial.begin(9600);
pinMode(ledpin, OUTPUT);
myservo.attach(9);
myservo.write(90); // set servo to mid-point
}void loop() {
//collects data from an analog sensor
val = analogRead(sensorpin); // reads the value of the sharp sensor
Serial.println(newval); // prints the value of the sensor to the serial monitor
delay(400);
newval= map(val, 1, 1000, 1, 10);
// wait for this much time before printing next value
if (newval<2) {
myservo.write(0);
digitalWrite(ledpin, LOW); // turn the LED on (LOW is the voltage level)
}
else
{
myservo.write(90);
digitalWrite(ledpin, HIGH); // turn the LED on (HIGH is the voltage level)
}
}
Final Product
Door lock
- Custom access for people whom you want/don’t want to give access to
- Motion sensor on the lock which opens by itself when a person appraches from inside.
- Motion sensor to detect any movement inside home. Recognizes cat, dog movement- so doesn’t notify then.
- Predicts common behavior of people and give best suggestions for access to ease out the experience.
- Coming soon: Open and close door with voice — Connects with Google assistant and Siri.
Door bell
- If someone rings the bell, owner gets notification on phone.
- Owners can video chat with the person on the door. And allow or deny access from the video call itself.
- Facial Recognition sensor that detects “access given” faces and opens the door automatically when they approach
- Coming soon: Encrypted cloud storage, Two factor authentication access. Facial recognition at door and finger print unlock at phone for a new person.
The App prototype
Home
The app landing page is simply a slider that locks and unlocks the device. It transitions from green (lock and safe) mode to red (unlock and unsafe) mode with another feedback of lock animating between the two states.
Access
The app lists all the faces coming to your home and categorize them based on people who have the access or who visit frequently. Once you give access to a face, the door bell will recognize the access and the lock will automatically open for that person.
History
View all video clips of people who came to your door today. And allow or deny their access for the future. If that person is on the door, you can video chat with him and give permissions while sitting remotely.
Product prototype
The lock fits right on top of existing deadbolt lock. So there are no efforts to put in the installation. And similarly, the door bell can also be simply mounted on the existing one.
As soon as the lock and bell are mounted, they can be connected to the internet using the home wifi. The products are ready to use.
The owner can scan his face on the app to give him owner access so that whenever he is arriving at the door step, the camera & facial recognition sensor detects the face and unlocks the door automatically. Owner can choose to disable facial recognition and keep the access only using the app, or if nothing else, by just the key itself.
Now, when a friend comes to home and rings the door bell, the owner gets notified on his phone and he can give/deny access to the friend remotely via the app. It’s that simple!
Moreover, owner can choose to give that friend’s face access, so that his friend can come in/go out anytime he wants, without troubling the owner every time.
When the door is opened using the app, app gives visual feedback by changing the color from green (safe & locked) to red (unsafe & unlocked) state. While the door lock responds with a visual feedback by glowing the light and turning the knob when opened by the app. When the light go out, the door can opened simply by rotating the knob.
Hope this exercise helps you understand the process and that you learned something new, as I did. Leave your thoughts and suggestions in the comments below.
Related links:
Framer prototype link:
Inspirations:
Technical Exploration
How to connect Siri with Framer prototype:
How to connect Arduino with HomeKit:
How to connect Arduino Uno to Wifi
How to send notification to phone from Arduino (I used Blink)
