Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Please more information about how it works #137

Closed
lnicalo opened this issue Jul 11, 2017 · 2 comments
Closed

Please more information about how it works #137

lnicalo opened this issue Jul 11, 2017 · 2 comments
Assignees
Labels
Module: Documents Indicates document issues Type: Question Indicates that an issue or pull request needs more information

Comments

@lnicalo
Copy link

lnicalo commented Jul 11, 2017

Thank you for this exciting project.

I am working on a project to develop a PoC self driving car. I am sure this platform can be useful. However I do not understand how the different modules guiding the car are integrated.

I would like to see more detailed documentation about how external information is collected, how objects are recognized to actually drive car.

Many thanks

@startcode
Copy link
Contributor

Thanks for your interest in Apollo.
The following document can answer part of your question.
From the Quit Start Guide

Apollo 1.0, also referred to as the Automatic GPS Waypoint Following, works in an enclosed venue such as a test track or parking lot. It accurately replays a trajectory and the speed of that trajectory that a human driver has traveled in an enclosed, flat area on solid ground.
At this stage of development, Apollo 1.0 cannot perceive obstacles in close proximity, drive on public roads, or drive in areas without GPS signals.

The future road map is disclosed in the Apollo Website http://apollo.auto/

@jinghaomiao jinghaomiao added the Type: Question Indicates that an issue or pull request needs more information label Jul 24, 2017
Capri2014 pushed a commit to Capri2014/apollo that referenced this issue Aug 5, 2017
Capri2014 pushed a commit to Capri2014/apollo that referenced this issue Aug 5, 2017
* add road boundary for perception

* fix unit test error

* [perception] add CNNSegmentation method (ApolloAuto#81)

* add unit test for lidar_process, and fix merge conflict

* fix merge conflict

* remove the comment of hdmap

* modify hdmap interface, and add unit test for hdmap_input

* Populate trajectory points directly in planner.

* fix a bug in prediction module

* add velodyne driver

* Fix planning lint errors.

* Updated evaluator and predictor managers

* [perception] clean up 'caffe.BUILD' and modify 'cnnseg/BUILD' (update). (ApolloAuto#82)

* [perception] add CNNSegmentation method

* [perception] clean up 'caffe.BUILD' and modify 'cnnseg/BUILD'

* add road boundary for perception

* fix unit test error

* [perception] add CNNSegmentation method (ApolloAuto#81)

* add unit test for lidar_process, and fix merge conflict

* fix merge conflict

* remove the comment of hdmap

* modify hdmap interface, and add unit test for hdmap_input

* Populate trajectory points directly in planner.

* fix a bug in prediction module

* add velodyne driver

* [perception] add CNNSegmentation method

* [perception] clean up 'caffe.BUILD' and modify 'cnnseg/BUILD'

* Modified container obtaining function name

* use config file in qp_spline_path_sampler. removed unused interface.

* uniform the naming of 'QP'

* fixed type comparison warning in rtk planner.

* added dp_road_graph_test

* prediction module lint fix

* added two util functions

* Cleaned up factories in prediction

* record planning input into debug pb.

* updated plot planning.

* Store reference line debug information only when enabled.

* updated plot planing with data lock

* Updated container names from string to adapter config message type

* record planning init point into debug.

* fix a bug in prediction module

* record replan info into debug pb.

* updated plot planning with map file arg

* render decisions in dreamview frontend

* Updated container manager to initialize with adapter config

* Minor code restructuring in planning (ApolloAuto#113)

* Added initialization for all managers

* Record planning total time latency stats into pb

* Add device permission when adding user.

* feature: fix some bugs in dp road graph (ApolloAuto#118)

* container/obstacles/obstacle.cc

* [perception] add 'use_all_grids_for_clustering' mode

* Populate path_optimizer path into debug pb.

* added vehicle in plot planning and added default map file

* moved unused functions in dp_st_cost to private.

* Combine Path and PathDebug.

* changed some config parameters for more centralized planning curve.

* implemented dp algorithm in dp_road_graph

* Added initializations in prediction framework

* updated plot planning due to updated pb

* added more dp_road_graph_test

* fix a bug in prediction module

* refactor path, discretized_path and frenet_frame_path

unify function names

* feature: add reference line on the road and get lane width functions (ApolloAuto#130)

* Add prediction to dreamview (frontend)

* fix gdb cannot debug problem in docker

reference https://stackoverflow.com/questions/34856092/gdb-does-not-hit-any-breakpoints-when-i-run-it-from-inside-docker-container

* modify some velodyne driver parameters in launch file (ApolloAuto#133)

* add velodyne driver

* modify some param

* fix path get_width = 0 problem (ApolloAuto#135)

the problem is because LaneWaypoint is not set in hdmap::Path.
fixed the broken unit tests

* purge functions that returns mutable vector (ApolloAuto#136)

* purge logic map related code (ApolloAuto#137)

* rename SetConfig to Init.

* modify perception config file (ApolloAuto#140)

* add unit test for lidar_process, and fix merge conflict

* fix merge conflict

* remove the comment of hdmap

* modify hdmap interface, and add unit test for hdmap_input

* modify perception config file

* enable planning log to stderr to make it easier to debug.

* disable stderr log for the latency is too worse.

* fix a bug that lane sequence prediction probability was always zero

* remove path.h in planning/common/path

* not a useful base class
* confusing with hdmap path.h

* Kimifly06 dev (ApolloAuto#139)

* push test

* push test

* add hm tracker

* update hm tracker code style

* feature: refine reference line projection logic

* fix a bug in prediction module, the problem was index out of range in lane segment

* remove invalid encoding char.

* simplify reference line smoother behavior

After change, the smoother will smooth any given reference line without truncate.
The behavior will be more transparent behavior.

* enable planning glog buffer.

* fix Hardware check build err after apply ESD lib (ApolloAuto#265)

* Fix esdcan build dependency.

* Make linear interpolation function template

* feature: add speed limit as well as utils  in qp spline st optimizer

* 	added num of points in smoother config proto. (ApolloAuto#154)

* formalized function names in dp_st_speed graph and cost.

* added num of points in smoother config proto.

* added speed info in plot planning.

* Added debug info

* use shared_ptr in hdmap/path to avoid future map memory recycle issue

* Added pseudo-inverse computation for non-square matrix

* Fixed prediction model proto

* Added some additional debug info

* refactor some logs in prediction module

* fix a small bug in prediction module

* Bug fix for non-square matrix inverse

* fix ros.BUILD

* bugfix: fix issues that s coordinate out of reference line range

* fix bug in dp poly path

* Added prediction publishing message header (ApolloAuto#167)

* Updated pedestrian predictor

* Added publish prediction header

* added lane_waypoints when empty in planning.

* Minor changes in prepartion for adding SimControl

* Fixed prediction lint error

* add processor time stats.

* Change routing request to routing result in adapter.conf

* turn off enable kf tracking by default, just for debuging on current rosbag with fake perception

* added init test for path.h/cc

* Add PrintIter helper to print containers.

* Fix prediction.js bugs

* Enabled prediction configuration in evaluator and predictor managers

* feature : add clear function for path data and speed data

* Fix lint errors.

* updated decision data

* HMI runtime_status_api returns meaningful message.

* update params of velodyne driver

* added feed by file feature for VehicleState class

preparation for more unit tests

* disable prediction input in planning by default.

* reorganize setup adapter code in base tests

* add visualizer lib

* Add helper function to print DebugStrings.

* added hdmap_line_path test.

* moving SpeedPoint to common/path_point.proto

* remove sl_point input in path sampler

* add speed plan into planning debug.

* added RoutingHelper to create Path from routing

* added path_test: hdmap_curvy_path test.

* added hdmap_circle_trajectory test.

*  added hdmap_jerky_path test.

* remove move constructor in pnc_map/path to allow copy

* migrate Point in map to pointENU

* let test build exit when first batch test failed.

* record optimizer speed data into debug.

* enable prediction in planning but not enforce prediction check yet.

* fix trajectory relative time bug in aggregation and protect trajectory stitch with correct flag.

* added TestSuite, hdmap_s_path

* added hdmap_path_get_smooth_point test

* Suppress logging from libcurl (ApolloAuto#203)

* added compute_lane_segments_from_points for path_test.

* fixed failed test.

* refactor lidar_process a little

* adding visualizer tool

* use shared pointer of LaneInfo within the prediction module

* [Perception] Fix shared_ptr bugs in cnn_segmentation.cc and merge fix (ApolloAuto#207)

* [perception] add 'use_all_grids_for_clustering' mode

* Add unit test for CNNSegmentation (libproto conflict issue should be fixed for Caffe)

* (1) Fix bugs of shared_ptr in cnn_segmentation.cpp; (2) Modify CNNSegmentation config; (3) Modify .cpp suffix to .cc

* Fix shared_ptr bugs in cnn_segmentation.cpp

* Modify cnnseg BUILD according the fix of ros.Build

* remove uni-test because of the libprotobuf conflict issue

* refactor tooling

* clear kf lane trackers once the obstacle is not on lane

* small change on log in prediction module

* added speed plot in plot planning

* added hdmap const ptr into path.

* changed planning code according to changes in path (with hdmap const ptr in path)

* changes in path_test.

* use translate point in free move prediction trajectory

* Updated model proto
@Zhenni17 Zhenni17 added the Module: Documents Indicates document issues label Jun 5, 2018
@Zhenni17
Copy link

Zhenni17 commented Jun 5, 2018

Hi @lnicalo, @startcode provided a good link for Apollo 1.0. The latest version for Apollo is 2.5, and here is a good place where you can start to read about:
https://github.com/ApolloAuto/apollo/blob/master/docs/quickstart/apollo_2_5_quick_start.md

If you'd like to learn more about each version of Apollo, here is a place that provides tons of info:
https://github.com/ApolloAuto/apollo/tree/master/docs/quickstart

If you have additional questions, feel free to re-open this issue. Closing the issue now.

@Zhenni17 Zhenni17 closed this as completed Jun 5, 2018
GowthamKudupudi pushed a commit to amd/apollo that referenced this issue Nov 17, 2022
[AMD][temporary] Introducing remap and duplicate kernels in image_data_operations
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Module: Documents Indicates document issues Type: Question Indicates that an issue or pull request needs more information
Projects
None yet
Development

No branches or pull requests

5 participants