lgsvl / autoware-data Goto Github PK
View Code? Open in Web Editor NEWLicense: Other
License: Other
Hello All,
Thanks for the simulator package its very useful one. I am using autoware quite a long time and recently started use lgsvl simulator with autoware.
I have followed the steps provided in here https://github.com/lgsvl/Autoware. Separately, Simulator and Autoware running successfully. But i am unable to test the autoware along with the simulator because of the pcd loading failed error.
The pcd files provided in the link https://github.com/lgsvl/autoware-data,
But when i load pcd files (autoware-data-master/data/map/pointcloud_map_sf_portion/) I am getting error, "failed /home/autoware/.autoware/data/map/pointcloud_map_sf_portion/DuckieSimulator_SF_Environment_00000.pcd"
COuld anyone please help me to fix this issue. Thanks in advance.
Hello,
I wonder how to generate PCD maps for NDT localization in Autoware if I have a batch of PCD files and their real poses. I think the maps you provide is generated by driving the car in the simulator and collect the LiDAR frames data. I am not sure how I can stitch such pcds files together.
Thank you very much.
Hi,
I'm using this repo to run autoware + lgsvl, following the start guide. When running the my_motion_planning.lauch file in the autoware runtime_manager, there was an error: could not find the vehicle_socket.launch. When changing the my_motion_planning.launch from
<!-- Vehicle Contorl -->
<include file="$(find runtime_manager)/scripts/vehicle_socket.launch"/>
to
<!-- Vehicle Contorl -->
<include file="$(find runtime_manager)/launch_files/vehicle_socket.launch"/>
the file could be properly found and the car drove correctly.
This file path correction work when considering autoware version 1.12.0. Not sure if it's kept for backwards compatibily, but reporting just in case.
Hi lgsvl,
I created a new vehicle and a new sensor layout, does lgsvl any recommand for how to calibration it?
thanks.
Hello everyone,
Could anyone please guide me, how to test autoware localization with lgsvl simulator.
Currently, I have followed all the steps mentioned in the link https://github.com/lgsvl/Autoware and
Until launch Rviz - the vector map and point cloud map display everything working fine. But next, i am unable to visualize the simulated LiDAR data from the simulator.
Could anyone please help me, how to connect real-time simulated lidar data with the autoware localization node.
Even though I have clicked "2D Nav Goal" in Rviz, then click a destination point and drag slightly in a feasible direction, i am unable to see any changes in rviz.
Regards,
Ajay
i have collected .pcap file using VLP32C and converted to a point cloud and .bag file but when i tried to localize the ros bag playing is not on the map. someone please help me. i can email my .pcd file and bag for someone interested to help and the tf i used is the one in autoware/src/.config
Hi,
The vector maps created for autoware doesn't work with "Autoware 1.13 -> op_planner"
Have you tried using these vector maps in op_planner?
Op_planner checks version (version 1 and version 2) for vector maps. When the vector maps data provided by lgsvl (autoware-data) is loaded, op_planner says it is version 2 vector map however when it reads the data it gives a segmentation error at the end of the "lane.csv" file. Because the last value of "FNID" in "lane.csv" doesn't exist in "node.csv" as a value of "NID". It is same for every vector map data provided by LGSVL in "lgsvl/autoware-data" repo.
OS : Ubuntu 18.04.4
Autoware Version: 1.13
Do you have any idea what the problem can be?
Thanks a lot in advance.
Mehmet Doğru
Ubuntu 18.04
LGSVL 2020.06
I am trying to create simple scenarios with ego vehicle following NPC driving straight ahead just like examples given default for Apollo. I am trying to use Autoware.AI ADS for ego vehicle. The problem is I am not able to localize ego vehicle in most of the maps. It does localize in non-API mode after trying few positions clicking on "2D Pose Estimate". But the same is not working uniformly with Autoware bridge connected in API mode.
Hello,
I want to try SingleLaneRoad map instead of autonomoustuff_parking_lot map. For this I need to change the variable "map_osm_file" in avp_core.launch.py script, "map_pcd_file" and "map_yaml_file" in the avp_sim.launch.py script. They have those values in the code:
map_osm_file = autonomoustuff_parking_lot.osm
map_pcd_file = autonomoustuff_parking_lot_lgsvl.pcd
map_yaml_file = autonomoustuff_parking_lot_lgsvl.yaml
I downloaded SingleLaneRoad.osm file from LGSVL simulator and SingleLaneRoad.pcd file from this repo. However, I couldn't find SingleLaneRoad_lgsvl.yaml file. How can I find it?
Hi all,
I have my own map an PCD map, and I want to draw vector map in my own map. What tools can I use?
Thanks.
Is calibration file the correct file?
in calibration file, I saw the Image Size. it was [640, 480].
but, lg simulator image size is FHD.
if i use this calibration file on the Autoware, then not working.
so, i changed the simulator image size from FHD to [640, 480].
then, algorithm is working, but lidar-camera data is not matched.
could I get a correct calibration file?
I am trying to use Autoware.Ai 1.14.0 with the SanFrancisco vector and pcd map. I am localizing the vehicle using the GPS odometry sensor available in the LGSVL.
This is what I am getting:
Since, I am using the ground truth localization from the GPS Odometry sensor. I believe the map is incorrect.
Thank you very much.
According to the docs, after I clicked the "Localization" in Autoware's GUI, there should be a time display shown on the right.
But It's not showing up in my case, same as the screen shot shown in above question.
Is this normal or did I miss something ?
Thank You
Originally posted by @ntutangyun in #2 (comment)
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.