https://github.com/a3ro-dev/sentinel-x
a autonomous robot control system using an ESP32 microcontroller. The robot features dual motor control, servo steering, and ultrasonic obstacle detection.
https://github.com/a3ro-dev/sentinel-x
Last synced: 10 days ago
JSON representation
a autonomous robot control system using an ESP32 microcontroller. The robot features dual motor control, servo steering, and ultrasonic obstacle detection.
- Host: GitHub
- URL: https://github.com/a3ro-dev/sentinel-x
- Owner: a3ro-dev
- License: mit
- Created: 2025-01-04T11:25:24.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-11-05T06:54:43.000Z (8 months ago)
- Last Synced: 2025-11-05T08:24:40.933Z (8 months ago)
- Homepage:
- Size: 29.3 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Sentinel-X
This project implements a autonomous robot control system using an ESP32 microcontroller. The robot features dual motor control, servo steering, and ultrasonic obstacle detection.
## Hardware Requirements
- ESP32 microcontroller
- 2 DC motors with H-bridge driver
- Servo motor
- HC-SR04 ultrasonic sensor
- Power supply
## Pin Configuration
```
Motors:
- IN1: GPIO18 (Motor 1)
- IN2: GPIO5
- IN3: GPIO15 (Motor 2)
- IN4: GPIO2
- EN1: GPIO19 (PWM)
- EN2: GPIO21 (PWM)
Servo:
- SERVO: GPIO4 (PWM)
Ultrasonic Sensor:
- TRIG: GPIO12
- ECHO: GPIO14
```
## Features
- Basic movement controls (forward, backward, left, right)
- **Enhanced Algorithm Features:**
- Exponential Moving Average (EMA) filtering for smoother distance measurements
- Median filtering for robust outlier rejection
- Weighted path scoring that prefers forward direction
- Path width estimation for better navigation decisions
- Adaptive safe distance based on current speed
- Proportional steering for smoother, more natural turns
- Predictive obstacle avoidance using approach rate calculation
- Adaptive servo sweep density based on obstacle proximity
- Intelligent obstacle detection and avoidance with multi-angle scanning
- Hardware initialization and validation
- Error handling and recovery with error threshold
- Speed control via PWM with differential turning
- Continuous servo scanning for improved obstacle detection
## Operation
The robot:
1. Initializes all hardware components (motors, servo, ultrasonic sensor)
2. Continuously sweeps the servo between 45° and 135° with adaptive step size:
- Slower, denser scanning (2° steps) when obstacles are close (< 50cm)
- Normal scanning (3° steps) at medium distances (50-100cm)
- Faster scanning (5° steps) when path is clear (> 100cm)
3. Takes 3 distance readings per measurement with advanced filtering:
- Uses median of readings for outlier rejection
- Applies Exponential Moving Average (EMA) for smoothing
- Validates readings within HC-SR04 range (2-400cm)
4. Calculates adaptive safe distance based on:
- Current robot speed (higher speed = larger safe distance)
- Approach rate (detects rapidly closing obstacles)
- Increases buffer distance when approaching obstacles quickly
5. If an obstacle is detected within adaptive safe distance:
- Stops immediately
- Backs up for 500ms
- Performs a three-angle scan (45°, 90°, 135°) to find the clearest path
- Calculates weighted scores for each path (prefers forward direction)
- Estimates path width from left and right measurements
- If no clear path is found (< 35cm in all directions), backs up for 1000ms
- Otherwise, applies proportional steering toward the best direction
6. Proportional steering provides smooth, natural turns instead of sharp binary turns
7. Uses the last valid distance reading if current measurement fails
## Error Handling
- The system performs 3 ultrasonic readings per measurement for accuracy
- Advanced filtering pipeline:
- Median filtering for robust outlier rejection
- Exponential Moving Average (EMA) smoothing for noise reduction
- Validates readings to ensure they're within HC-SR04 range (2-400cm)
- Hardware initialization checks all components before operation
- Main loop includes error recovery with a maximum error threshold (5 errors)
- System will safely stop if too many consecutive errors occur
- Emergency turn performed on error recovery
- Graceful degradation: uses last valid reading if current measurement fails
## Usage
Upload `ota/car.py` to an ESP32 with MicroPython installed. The code will run automatically on boot.
The robot will autonomously navigate while avoiding obstacles. No additional user input is required during operation.