PatrolBot

This Halbot inspired robot makes use of matlab processing to detect voice commands and will picture any suspicious characters with motion detection, uploading the culprits photo to the cloud.

Inspiration

Motivated by personal security and the new IoT era, what better method to improve our safety than to have a moving pair of eyes to watch your back. Away from traditional easily-evaded mounted cameras or noise detectors, our new invention provides all the functionality of both, and more.

How we build our project and technical specs:

Parts:

• Camera: Arducam Mini Module Camera Shield 5MP Plus OV5642 Camera Module for Arduino UNO Mega2560
• Board
• Microphone: MAX9814 Electret Microphone Amplifier
• Halbot
• Ping sensor: kuman 5pcs Hc-sr04 Ultrasonic Distance Measuring Sensor Module
• Bluetooth: HC-05 Bluetooth serial module
• 3 Arduino UNO
• 2 Diodes: 1N4001 Diode
• 2 MOSFETS: IRLB8721 N-Ch MOSFET
• LED
• Buzzer

We began our project once we had obtained all necessary parts. These parts included an Arduino camera,, microphone, Halbot, ping sensor, bluetooth modules, and Arduino UNOs. We created multiple independent systems, one for each type of hardware we had, to understand and test each component. Each of these isolated components had their unique code in arduino and some in MatLab. Once we had a basic grasp of our components, we began assembly of the complete product, conglomerated each isolated component onto a few breadboards and arunios. The system would have two main forms of inputs; the voice recognition and sensors. From our computers, we had a voice recognition software in MatLab, capable of interpreting simple commands into something the Arduino could interpret through serial communication. This would be our main form of controlling the movement of the PatrolBot itself. Mounted on the actual Halbot, the ping sensor, microphone, and camera would act as our eyes an ears. If the Halbot were to detect activity in its proximity either through its ping sensor or microphone, the Halbot would alert the user by sending a signal to the bluetooth module mounted on the Halbot itself to a bluetooth device the user would be carrying. For our personal device, we used a LED and buzzer wired to the receiving bluetooth module. 

Challenges and Overcoming Them:

Challenges began appearing immediately after trying to assemble the PatrolBot. When we had the isolated components, it was easy to run and test each component. However, getting these components to communicate with each other was another issue in itself. The most difficult issue was the voice recognition. The voice recognition had a low success rate at best when we began testing. Be eventually resolved this by reducing the amount of input MatLab would have to store and process.

What We Learned:

Most of the components of this project involved labs we had done before. Regarding the build process, we had to learn MatLab to Arduino communication and apply that to the PatrolBot.

What’s Next for the Project:

In the future, we would like to add a patrol function to PatrolBot. The objective would be to give PatrolBot an autonomous function. Also, we would like to integrate a WiFi system that allows a greater distance for communication between the bot and the user. Currently, the live camera must be attached to the computer via a USB to arduino camera. If possible, we could be able to transmit the data from the camera via WiFi. We would also need to improve the voice recognition software accuracy and functionality.

Built With

• arduino
• wifi