{"id":17632705,"url":"https://github.com/fatlipp/toyslam","last_synced_at":"2026-04-20T03:02:52.634Z","repository":{"id":250599816,"uuid":"834861083","full_name":"fatlipp/ToySlam","owner":"fatlipp","description":"SLAM implementation from scratch w/o external graph optimization libs","archived":false,"fork":false,"pushed_at":"2024-09-07T14:43:36.000Z","size":256,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-03-30T03:31:48.598Z","etag":null,"topics":["cuda","gpu","lidar-slam","mapping","odometry","robotics","slam"],"latest_commit_sha":null,"homepage":"https://medium.com/@fatlip/graph-based-slam-basics-f84501525f24","language":"C++","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/fatlipp.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-07-28T15:30:27.000Z","updated_at":"2024-09-07T14:43:38.000Z","dependencies_parsed_at":"2024-07-28T19:07:48.637Z","dependency_job_id":"f0e5fe2b-cbef-42db-8c3a-e28048f1d475","html_url":"https://github.com/fatlipp/ToySlam","commit_stats":null,"previous_names":["fatlipp/toyslam"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/fatlipp/ToySlam","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fatlipp%2FToySlam","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fatlipp%2FToySlam/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fatlipp%2FToySlam/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fatlipp%2FToySlam/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/fatlipp","download_url":"https://codeload.github.com/fatlipp/ToySlam/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fatlipp%2FToySlam/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32031070,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-20T00:18:06.643Z","status":"online","status_checked_at":"2026-04-20T02:00:06.527Z","response_time":94,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cuda","gpu","lidar-slam","mapping","odometry","robotics","slam"],"created_at":"2024-10-23T01:45:12.868Z","updated_at":"2026-04-20T03:02:52.617Z","avatar_url":"https://github.com/fatlipp.png","language":"C++","funding_links":[],"categories":[],"sub_categories":[],"readme":"## Simple SLAM Implementation\nThis project is a basic implementation of SLAM using a simulated 2D LiDAR. The goal is to demonstrate the fundamental principles of SLAM, including robot pose estimation and optimization using graph.\n\n![SLAM](/assets/SLAM.png)\n\n### Features\n- 2D LiDAR Simulation: A virtual LiDAR sensor is simulated, providing distance measurements to surrounding objects. This data is used to map the environment and estimate the robot's pose.\n - Robot Pose Optimization: The robot's pose (position and orientation) is optimized using noisy landmarks and noisy robot positions.\n - Remote CPP optimizer:\n   - 1. Build `cpp` folder:\n     Example of using conan:\n        * conan install . --output-folder=./build -s compiler.cppstd=gnu20 -s compiler.version=13 --build=missing\n        * cmake --preset conan-release -DWITH_CUDA=OFF [ON/OFF]\n        * cmake --build --preset conan-release\n   - 2. Run: graph_optimizer HOST PORT ITERATIONS PIPELINE SOLVER (./bin/graph_optimizer \"127.0.0.1\" \"8888\" \"50\" cpu eigen)\n        - ITERATIONS - [int value \u003e= 1]\n        - PIPELINE: [cpu/gpu]\n        - SOLVER: [cuda/eigen] (if `PIPELINE == GPU` =\u003e `SOLVER = CUDA`)\n   - 3. Run `python3 python/slam_main.py`\n\n### Requirements\n - numpy: for matrix operations\n\n - CPP (optional, for remote optimization):\n    * Eigen - for matrix operations\n    * Conan 2 (optional) - packet manager\n    * CUDA (optional)\n\n### Pipeline\nThere is 4 different pipelines:\n - Python UI + Python optimizer\n - Python UI + CPP optimizer\n - Python UI + CPP optimizer (CUDA matrix solver)\n - Python UI + Full CUDA optimizer\n\n### Key concepts\n - State Representation: The robot's state is represented by 3x3 matrix:\n**[R t]**\n**[0 1]**\n\n - Odometry:\n**[R t]**\n**[0 1]**\n\n - Graph2d: consists of robot poses in 2d and relative landmarks observations\n - OptGraph: consists of robot poses in 2d and relative landmarks observations, that used for optimization\n - 2 types of edges: [ODOM, LM]\n\n### Simplification\nEach point in the environment has its own ID, which is used to match landmarks\n\n### Further development\n - Adding More Edge and Vertex types (2d, 3d, BA, Virtual Meas.)\n - Using a map for navigation\n - Performance optimization\n - CUDA opt improvements","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffatlipp%2Ftoyslam","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ffatlipp%2Ftoyslam","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffatlipp%2Ftoyslam/lists"}