Obstacle avoiding robot | MoveBase using Lidar | Webots ROS | Robotic Software PicoDegree | Part 6

Wish to get into the shoes of a Robotics Software Engineer and see the complete cycle of mobile robot development. Also learn and implement robotics concepts using ROS with a great simulator named Webots. Then you are at the right place. Soft_illusion Channel is back with a new video..!! (A channel that aims to help the robotics community).
#Obstacle_avoidance #Navigation #Move_base #ROS
00:15 Introduction
02:09 Details of Obstacle avoidance
03:01 Object avoidance
05:50 Glimpse of Obstacle avoidance
06:10 Setup the repo
08:55 Remap scan topic
09:51 Update local cost map to include laser data
15:39 Recovery behavior MoveBase
16:30 Clear costmap service
RBR50 2021 Mobile robot innovation winners :
mobilerobotgui...
2 parts in obstacle avoidance:
Obstacle detection and obstacle avoidance.
Obstacle detection: This can be done by 2D/3D camera or Lidar.
Obstacle avoidance: We can start with basic algorithms like Bug
jwcn-eurasipjo...
or Implement Potential fields with graph search Algorithms.
Different algorithm techniques:
Bug2 Algorithm - Going around the obstacle on the way.
Artificial Potential Field - Treating obstacle as repel and goal as attracting object.
Collision Cone - Treating obstacle as a circle.
Fuzzy Logic - Taking in multiple sensor input and doing fuzzification of the data.
Vector Field Histogram - converting 2d obstacle space to a 1d histogram using a lidar.
Neural Network - Human like method where is trains itself and get better.
Main 2 changes in the code to add obstacle avoidance :
1. Remap topic file:
If you see this topic which we need to add in costmap changes every time we launch the project because of the namespace of webots. So we will remap the topic name to a constant string as “laserscan”
Steps involved in that are :
a. We subscribe to the model name to get the namespace name
b. We subscribe to laser topic to get the data
c. We publish same data to a “/laserscan” topic.
2. Adding obstacle layer in the 2d local costmap using costmap_2d::ObstacleLayer plugin if we have a 3d application we can use costmap_2d::VoxelLayer instead. We also give the rights for sensor data to add and remove obstacle cost.
Finally, In the end, we see the implementation of recovery behavior. We use the move base default settings and also we see how to clear the costmap with an inbuilt service in move base which is known as rosservice call /move_base/clear_costmap {}

[@IshitaShahi]

You guys make it so interesting with animation. Kudos team 👏👏

[@coolrobotics]

Thank you Ishita 😇

[@markusbuchholz3518]

I would like to thank you for next inspiring video. This channel is remarkable and inspiring. Great you promote ROS and express beautifully the essence of robotics. I keep fingers for your growth as you can imagine. Have a nice day!

[@coolrobotics]

Thanks Markus ….
For the great support 🙏🏼

[@PranshuTople]

Very well explained in detail!!! Just one suggestion - increase the pace of your video.
If you are a viewer, I would recommend to watch this video in 2x speed

[@coolrobotics]

Sure …
Noted ✌🏽

[@krisshcool6381]

@9:08 when i do rostopic echo /model_name. I didnt get any output. So what should i keep in the code for model_name ? TIA

[@smatef]

Can you do the same with using ultrasonic instead of LiDAR?

[@coolrobotics]

It can be done but ,
It will be a bit difficult to gauge the size of obstacle

[@will-le3kl]

hello, I'm new to webots and ros, Can lidar be displayed like an image， depth image ?

[@coolrobotics]

You can use 3D camera for this purpose to get 3D point cloud

[@casper5314]

How do you enable PointCloud on the lidar?

[@yusufcicek4596]

more subtitle please.

[@rheymint1836]

Good day! Thank you for this video. Can I ask for the code for this? Thank you sp much!

[@coolrobotics]

You should be able to see this code on GitHub.
Link should be there is the description

[@abishakinbanathan]

how to communicate with physical bot ?