The Forest Security System is a smart monitoring project built with Quarky and PictoBlox. The system combines environmental sensing, motion detection, AI-based recognition and IoT dashboard monitoring to help identify potentially dangerous activity in a forest or protected area.

The PictoBlox program continuously reads temperature, motion and smoke/gas-related inputs, checks for a person using computer vision, and uses an audio/ML recognition model for a chainsaw/hand-saw type alert. Important events are sent to Adafruit IO so they can be monitored remotely.

Main Components

1. Quarky controller / Expansion Board
2. DHT sensor
3. PIR motion sensor
4. Smoke/gas or LDR-type analog sensor
5. Servo motor connected through Quarky expansion hardware on servo channel 5
6. Computer with PictoBlox and webcam for AI/object detection
7. Speaker/buzzer output supported by Quarky
8. Wi-Fi / Internet connection for Adafruit IO
9. Adafruit IO account and feeds

1. Power the Quarky and connect the required sensors.
2. Start PictoBlox and open the Forest Security System project.
3. Initialize the Quarky expansion board and set the display brightness.
4. Connect PictoBlox to Adafruit IO using the account credentials.
5. Inside the main forever loop, read the DHT sensor and store the temperature in the TEMPRATURE variable.
6. Send the temperature value to the Adafruit IO feed named Temperature.
7. Read the PIR sensor. If motion is detected, send 1 to the flame feed, play a C4 warning note, display an alert pattern, and then clear the display. When motion is not detected, send 0.
8. Read the analog smoke/gas input. If the sensor reports detection, send 1 to the Smoke feed and play a warning tone. Otherwise send 0.
9. Run the PictoBlox ML recognition routine and check the Chain Saw confidence. The project uses a threshold greater than 0.95 before sending a chainsaw alert.
10. Turn on the webcam video feed and run object detection. If a person is detected, send 1 to the Person feed and upload the detected object count to the No. feed. Otherwise send 0 to both.
11. A separate ML condition checks class3 and reports it through the HAND SAW feed.

A second script initializes an Angle variable and sweeps servo channel 5 from 0° to 180° and back, creating a continuous scanning motion

1. Audio/ML recognition: the project checks the confidence of the 'Chain Saw' class and requires confidence above 0.95 before generating the alert.
2. Object detection: PictoBlox analyses the webcam image and checks whether a 'person' is detected.
3. People count: when a person is detected, the project sends the detected object count to the Adafruit IO 'No.' feed.
4. A separate ML class3 condition is used for the 'HAND SAW' alert feed.
5. These AI features allow the system to combine sensor data with visual and audio information instead of relying on only one sensor.

1. Add GPS/location reporting so an alert includes the approximate monitored-zone location.
2. Add a real-time alert service such as email or mobile notifications.
3. Store historical temperature, smoke and intrusion data for trend analysis.
4. Add solar power and battery monitoring for remote deployment.
5. Add multiple sensor nodes and compare readings from different forest zones.
6. Add a local OLED/LCD status screen with clear alert labels.

This Forest Security System demonstrates how Quarky, PictoBlox, sensors, AI and IoT can be combined into a single monitoring platform. The project monitors temperature and environmental/security inputs, uses AI to recognize people and sound-related threats, and publishes important information to Adafruit IO. The same architecture can be expanded for wildlife protection, campus security, agricultural monitoring or other remote-area safety applications