Questions Regarding Map Construction and Navigation Approach

[JinyeonKim]

Hello,

Thank you for your impressive work. We're currently exploring the possibility of applying your research to our environment, and I have a few questions regarding the map generation process.

We're particularly having difficulty with the step of converting a given environment into a usable map. Could you share more details on how you approached the data collection for this purpose? Any specific methods, tools, or pipelines you used would be very helpful.

Also, based on my understanding, the navigation process doesn't involve real-time environmental sensing and dynamic collision avoidance. Instead, it seems to rely solely on a pre-constructed map for navigation. Could you please confirm if this is correct?

Looking forward to your clarification.

Thank you!

[ShahRutav]

Hi @JinyeonKim,

Thanks for reaching out.

Could you share more details on how you approached the data collection for this purpose? Any specific methods, tools, or pipelines you used would be very helpful.

By data collection, do you mean the 2D map generation? We used the ros gmapping package for 2D map generation.

the navigation process doesn't involve real-time environmental sensing and dynamic collision avoidance. Instead, it seems to rely solely on a pre-constructed map for navigation. Could you please confirm if this is correct?

I used a global planner with a local planner in the GoToLandmark skill for navigation. Both local and global planners react to dynamic obstacles using onboard sensors. I must mention that not all skills are reactive to dynamic obstacles.

Hope this helps!