85b192530d
* Initial implementation of t.m. performance script * Remove redundant getLocation() calls * Demo for inter-client communication via Carla server * WIP: To do: For client usage needed to be changed. * Instead or client instance episodeProxy is passed to TM * Instead or client instance episodeProxy is passed to TM * parmeter improvements (walkers,cars,signs,lights) * adding section id to map setup * fix manual_control to reload car with autopilot on * Instead of client instance episodeProxy is passed to TM. * WIP: Only added vehicle register via RPC server / client call in TM to local instance from remote instance. Similarly other APIs we need to added. * Added all TM APIs * carla client now can provide TM instance if required. * While getting TM instance if server given data is not valid new TM instace is created. Need to decide what to do with earlier registed vehicle with previous TM. * Updated with HealthCheckRemoteTM() function to check TM server status instead of ResetAllTrafficLights() for any new user client. * Dynamic port selection added to TM server instace (in case of multiple TM serve run, it required). * Fixed change map error * If registered TM at carla server is not present then remote TM should be freed and new local TM instance is created. * Created Process for Local TM. It cheaks with total registered vehicles in it, if no register vehicles for 5 sec, it closes it self. In python spawn_npc.py Client Timeout set to 10 sec. as creating Process is system dependent call and takes time. User has to unregister vehicle before destroying it (if activated with TM). Also neet to check TM unregister function to remove vehicles from register list. * Fix for collision ignore bug * Update TM exit function check to compare registered vehicles with world actors that if any valid vehicle present cointinue to run the TM else stop. * Print removal * Splited work of main for loop in ApplyBatchCommandsSync * WIP! Trying to get client directly * WIP! Trying to access episode properly without getting it from TM ctr * Changes to BatchControlStage for Sync Mode Changes to Parameters for Synch Mode * Added port support for TM. Multiclient MultiTM * Added additions for RPC Synchronous Tick to TM. Error handeling and code organization. * Exposed API for Sync Mode * TMServer notifies Server that it is gonna be destroyed * Exposed Python API for Sync Mode * Add TM as separate process and synchronous tick calls * SetSynchronousModeTimeOutInMiliSecond method added * TM shutsdown and informs to the connected clients * WIP! Map change issues again * Updated Traffic Manager for: 1) Multi-client, Multi-TM (as required) can be reistered to server. 2) TM is not a PROCESS separated from the user-client. 3) TM will exit if user-client exit. 4) If server closed, TM also get closed by catching runtime err. * Server side changes ... * Fixed possible stall on TM destruction * fixing collision stage + cybertruck not safe * Merge branch 'master' into 'soumyadeep/traffic_manager' * WIP! Disconnection of server has to be properly handled by clients * format update * Fix bug unsignalized junctions * # WARNING: head commit changed in the meantime Merge branch 'doterop/traffic_manager' into soumyadeep/traffic_manager Updates for Syncronized Tick. * Updated runtime exception in Episode for smooth exit if TN server closed for any other user client. * Better Exception handeling * Merged with jackbart94/tm_reduce_getloc_calls * Merged soumyadeep/traffic_manager. Removed World getter and recovered Episode as a parameter of TM ctr. Fixed syntax errors. * Initial implementation of t.m. performance script * Remove redundant getLocation() calls * Demo for inter-client communication via Carla server * WIP: To do: For client usage needed to be changed. * Instead or client instance episodeProxy is passed to TM * Instead or client instance episodeProxy is passed to TM * parmeter improvements (walkers,cars,signs,lights) * fix manual_control to reload car with autopilot on * Instead of client instance episodeProxy is passed to TM. * WIP: Only added vehicle register via RPC server / client call in TM to local instance from remote instance. Similarly other APIs we need to added. * Added all TM APIs * carla client now can provide TM instance if required. * While getting TM instance if server given data is not valid new TM instace is created. Need to decide what to do with earlier registed vehicle with previous TM. * Updated with HealthCheckRemoteTM() function to check TM server status instead of ResetAllTrafficLights() for any new user client. * Dynamic port selection added to TM server instace (in case of multiple TM serve run, it required). * Fixed change map error * If registered TM at carla server is not present then remote TM should be freed and new local TM instance is created. * Created Process for Local TM. It cheaks with total registered vehicles in it, if no register vehicles for 5 sec, it closes it self. In python spawn_npc.py Client Timeout set to 10 sec. as creating Process is system dependent call and takes time. User has to unregister vehicle before destroying it (if activated with TM). Also neet to check TM unregister function to remove vehicles from register list. * Fix for collision ignore bug * Update TM exit function check to compare registered vehicles with world actors that if any valid vehicle present cointinue to run the TM else stop. * Splited work of main for loop in ApplyBatchCommandsSync * WIP! Trying to get client directly * WIP! Trying to access episode properly without getting it from TM ctr * Changes to BatchControlStage for Sync Mode Changes to Parameters for Synch Mode * Added port support for TM. Multiclient MultiTM * Added additions for RPC Synchronous Tick to TM. Error handeling and code organization. * Exposed API for Sync Mode * TMServer notifies Server that it is gonna be destroyed * Exposed Python API for Sync Mode * Add TM as separate process and synchronous tick calls * SetSynchronousModeTimeOutInMiliSecond method added * TM shutsdown and informs to the connected clients * WIP! Map change issues again * Updated Traffic Manager for: 1) Multi-client, Multi-TM (as required) can be reistered to server. 2) TM is not a PROCESS separated from the user-client. 3) TM will exit if user-client exit. 4) If server closed, TM also get closed by catching runtime err. * Fixed possible stall on TM destruction * fixing collision stage + cybertruck not safe * Merge branch 'master' into 'soumyadeep/traffic_manager' * WIP! Disconnection of server has to be properly handled by clients * format update * Fix bug unsignalized junctions * # WARNING: head commit changed in the meantime Merge branch 'doterop/traffic_manager' into soumyadeep/traffic_manager Updates for Syncronized Tick. * Updated runtime exception in Episode for smooth exit if TN server closed for any other user client. * Better Exception handeling * Merged with jackbart94/tm_reduce_getloc_calls * Merged soumyadeep/traffic_manager. Removed World getter and recovered Episode as a parameter of TM ctr. Fixed syntax errors. * Rebased with master * Changes after rebase * Solving tab errors * Updated Changelog * Removing Destroyed vehicle's from TM Server * Fixed uint compatibility with Windows * Merged soumyadeep/traffic_manager. Discarded destroyed actors * -Removed unnecessary files * restoring docs from rebase * Fix windows compilation * refactoring stage-related code * more code refactoring * When map change the simulation doesn't throw exception anymore * Fixed incorrect episode * Fixed map change and TM remote detached thread destruction * Syntax and comments fixes * Missed change on previous commit * Fixed compile minor compile issue * Cleaned and fixed some issues after merge * fix to sync localization bugs modified PID parameters revamping spawn_npc * deleted tm_spawn_npc * fixes spawn error in sync mode * Redoing TM sync logic * finished performance benchmark for tm * deprecated wrapped methods: register_vehicle unregister_vehicle * New TM management * Fixed sync mode on TM * Cleaned TM of prints and unussed functions * collision stage checks for (0,0,0) to ignore. in memory map has an # between keys to avoid possible mixup. fixed spawn_npc with new sync mode * changelog * added more connection retries * fixed changelog + comments (see reviewable) * Moved socket include's to single header * Added missing line at the end of the file * Fixed syntax errors * final commit * Minor correction in comment * update copyright year to 2020 + removed break * restoring unwanted changes * patch for smoke test error * Removed "todo" for pylint Co-authored-by: Jacopo Bartiromo <32928804+jackbart94@users.noreply.github.com> Co-authored-by: Praveen Kumar <35625166+pravinblaze@users.noreply.github.com> Co-authored-by: Soumyadeep <soumyadeep.dhar@kpit.com> Co-authored-by: joel-mb <joel.moriana@gmail.com> Co-authored-by: Sekhar Barua <58979936+sekhar2912@users.noreply.github.com> Co-authored-by: bernat <bernatx@gmail.com> Co-authored-by: Marc Garcia Puig <marcgpuig@gmail.com> |
||
---|---|---|
.github | ||
Docs | ||
Examples/CppClient | ||
Import | ||
LibCarla | ||
PythonAPI | ||
Unreal/CarlaUE4 | ||
Util | ||
.gitattributes | ||
.gitignore | ||
.pep8 | ||
.readthedocs.yml | ||
.travis.yml | ||
CHANGELOG.md | ||
CMakeLists.txt | ||
Doxyfile | ||
Jenkinsfile | ||
LICENSE | ||
Makefile | ||
README.md | ||
Update.sh | ||
mkdocs.yml | ||
wheel_config.ini |
README.md
CARLA Simulator
CARLA is an open-source simulator for autonomous driving research. CARLA has been developed from the ground up to support development, training, and validation of autonomous driving systems. In addition to open-source code and protocols, CARLA provides open digital assets (urban layouts, buildings, vehicles) that were created for this purpose and can be used freely. The simulation platform supports flexible specification of sensor suites and environmental conditions.
If you want to benchmark your model in the same conditions as in our CoRL’17 paper, check out Benchmarking.
CARLA Ecosystem
Repositories associated to the CARLA simulation platform:
- Scenario_Runner: Engine to execute traffic scenarios in CARLA 0.9.X
- ROS-bridge: Interface to connect CARLA 0.9.X to ROS
- Driving-benchmarks: Benchmark tools for Autonomous Driving tasks
- Conditional Imitation-Learning: Training and testing Conditional Imitation Learning models in CARLA
- AutoWare AV stack: Bridge to connect AutoWare AV stack to CARLA
- Reinforcement-Learning: Code for running Conditional Reinforcement Learning models in CARLA
- Map Editor: Standalone GUI application to enhance RoadRunner maps with traffic lights and traffic signs information
Like what you see? Star us on GitHub to support the project!
Paper
If you use CARLA, please cite our CoRL’17 paper.
CARLA: An Open Urban Driving Simulator
Alexey Dosovitskiy, German Ros,
Felipe Codevilla, Antonio Lopez, Vladlen Koltun; PMLR 78:1-16
[PDF]
[talk]
@inproceedings{Dosovitskiy17,
title = {{CARLA}: {An} Open Urban Driving Simulator},
author = {Alexey Dosovitskiy and German Ros and Felipe Codevilla and Antonio Lopez and Vladlen Koltun},
booktitle = {Proceedings of the 1st Annual Conference on Robot Learning},
pages = {1--16},
year = {2017}
}
Building CARLA
Use git clone
or download the project from this page. Note that the master
branch contains the latest fixes and features, for the latest stable code may be
best to switch to the stable
branch.
Then follow the instruction at How to build on Linux or How to build on Windows.
Unfortunately we don't have official instructions to build on Mac yet, please check the progress at issue #150.
Contributing
Please take a look at our Contribution guidelines.
F.A.Q.
If you run into problems, check our FAQ.
License
CARLA specific code is distributed under MIT License.
CARLA specific assets are distributed under CC-BY License.
The ad-rss-lib library compiled and linked by the RSS Integration build variant introduces LGPL-2.1-only License.
Note that UE4 itself follows its own license terms.