robot car outdoor autonomous driving with GPS/liDAR
I have built a robot car with odom/lindar/camera etc. It runs very well indoor with a map built by gmapping. Now I am thinking a real world autonomous driving application. Suppose I want my car to go by itself from my house to my friend's house, or to drive autonomously along the concrete bycicle road in the park, what is the general approach I should take? I do not think gmapping will work here.
Maybe I can do this:
- Get some map like google map, or maybe "high definition map" from Here.
- Install a GPS module on my car
- Use google map and google map API to plan a path;
- Before driving the car, some how collect some lidar scan data along the path.
- When the car is driving, use GPS and google/Here map to find where it is and then use the saved lidar scan data to further locate itself accurately.
- Adjust speed/steering based on the true accurate locations.
I do not know if these steps are reasonable or feasible. What is the correct technical approach to drive the robot car outdoor?
Asked by AutoCar on 2018-12-20 00:46:50 UTC
Answers
At first beginning I want to write down some stuff but turned out introducing you baidu apollo's dev center course conducted by Sebastian Thrun, this is free and quite brief.
At least you could get some basic conceptions about autonomous driving and I think you could move to this from navigation package seamlessly.
Here it is.
http://apollo.auto/devcenter/devcenter.html
Asked by tianb03 on 2018-12-24 22:12:05 UTC
Comments