Sahayak - An Autonomous Covid Aid-Bot

Python Requirements

pip install -r requirement.txt

Required ROS Packages

xargs sudo apt install < ros_requirements.txt

CAD Model of Sahayak

• URDF and mesh files of the SolidWorks Assembly are generated using the sw_urdf_exporter.
• After installing the exporter, the SW2URDF add-in gets appended in the add-ins list of SolidWorks. Further detailed instructions are documented here.

Result

ROS Controls:

Teleoperation using Keyboard inputs

For teleoperation of Sahayak run the below command in terminal:

roslaunch sahayak_control teleop.launch

Visual Odometry:

Implemented Visual Odometry pipeline using the following

roslaunch visual_odom launch_all.launch
rosrun visual_odom optical_pnp.py 

• 3D to 2D Motion Estimation Algorithm Result Video

Mapping:

GMapping

GMapping is a ROS Package used creating a 2D occupancy grid map of an enviroment using a 2D LiDAR, tf and Odometry data.

To map the default enviroment using GMapping run the below commands in seperate terminals:

roslaunch sahayak_mapping gmap-launch_all.launch

roslaunch sahayak_navigation scan_matcher.launch

Hector SLAM

Hector SLAM is a ROS package which is used for creating a 2D occupancy grid map of an enviroment using a 2D LiDAR, tf data and Odometry data (Optional).

RTAB (Real Time Appearance Based) Map:

RTAB-Map is a ROS Package which uses a RGB-D camera to generate a 3D map of an enviroment.

To map the default enviroment using RTAB Map run:

roslaunch sahayak_mapping rtab-mapping.launch

ROS Navigation Stack:

AMCL

Sahayak was localised in a map using Adaptive Monte Carlo Localisation (AMCL) which uses adaptive particle filters.

Path Planning

Path Planning involves finding a lowest cost path based on 2D Local and Global Costmaps.

ROS Packages using in Path Planning:

• AMCL
• costmap_2d
• dwa_local_planner
• global_planner

Run roslaunch sahayak launch_all.launch in a terminal to launch Path Planning launch file. Use 2D Nav Goal in RVIZ to give a goal to Sahayak.