In this project, a simulation of a disaster environment was implemented using a TurtleBot3 hardware. The robot was used to survey the area by mapping its trajectory also the area and identify potential dangers and victims. The project demonstrated the efficiency of using a robot to map an area and identify hazards and injured people before sending human responders into the field. The performance of the mover_base and explore_lite packages were modified using ROS and Python in order to improve the robot's capabilities in terms of reliability and precision.
The purpose the this project is to simulate a reconnaissance mission after a disaster This could be a burned down building, hurricane, earthquake, etc. To reduce the risk to first responders, robots would be deployed in dangerous environments to survey and report back To model this scenario: turtlebot -> mobile platform apriltags -> victims closed room with obstacles -> dangerous environment
There are 2 types of Turtlebots:
- Burger.
- Waffle. We are using turtlebot burger for our application for its better maneuverability, operation time and lightweightedness as compared to turtlebot waffle.
General Specifications : Size : 138178192(mm) Maximum translational velocity: 0.22 m/s Maximum rotational velocity : 2.84 rad/s
Technical Specifications: Raspberry pi 3 (Single board Computer) 1024(1GB) RAM 64 bit Quad Core Arm Cortex - A53 Wifi : Dual band 2.4 & 5 GHz Power Supply 5V. OpenCR (Micro Controller Unit) 32 bit ARM Cortex -M7 with FPU Power Supply 5v
Actuator (XL430 - W250) Power supply 6.5-12V (11.1V recom). No load Speed with recom power 57 rev/s. Uses PID control Algorithm (Feedback : Position, Velocity, Load, Real Time tick, Trajectory, Temperature, Input Voltage, etc) Laser Distance Sensor (360 LDS -01) 2D 360 Deg laser scanner. Power Supply 5V. Ambient light resistance <= 10,000 luminescence
Camera - (Raspberry pi cam v2.1) Sensor IMX219 - 8MP Horizontal FoV - 62.2Deg. Vertical FoV - 48.8 Deg. IMU 3-Axis Gyroscope 3-Axis Accelerometer Battery Lithium polymer 11.1V 1800mAh / 19.98Wh 5C
Prior to starting to solve the problem: OS installation Network configuration Camera and Lidar nodes Camera Calibration Starting point for our solution Explore_lite Apriltag_detection
Found intrinsics of the camera using Kalibr Central Pixel, focal length Distortion coefficients Low reprojection error Less than a pixel in both axes
