Building Bridges: Leveraging C++ and ROS for Simulators, Sensor Data and Algorithms
Talkat CppCon 2023 · on Thursday, 2023-10-05· at 16:45· Cottonwood 8/9, 6700 North Gaylord Rockies Boulevard·
summary
Describes a concrete C++/ROS 2 implementation: a simulator-to-ROS communication bridge that converts CARLA sensor outputs into ROS messages and back to control, plus an execution-management framework (an Apex Grace SDK component) that enforces fixed‑order execution coordinator semantics. Claims deterministic, reproducible simulation loops by freezing the application to a single user thread, bounding memory and avoiding runtime allocs, and integrating CARLA with ROS 2 to produce consistent outputs across runs.
AI-written summary · about
chapters (20)
- 0:00Simulation and C++ with ROS
- 1:13Simulation in robotics
- 3:23ROS in robotics
- 6:33ROS communication basics
- 9:01Simulator-ROS connection
- 10:48Simulator-ROS bridge
- 12:38Determinism in simulation
- 13:41Determinism defined
- 15:48Data determinism pillars
- 17:53Benefits of deterministic sim
- 21:03ROS determinism limits
- 22:31Overcoming ROS determinism limits
- 24:17Determinism in simulation
- 26:04Key concepts for deterministic sim
- 29:57Apex Grace determinism
- 31:05Executor management framework
- 38:57Fixed order execution
- 44:38Fixing execution order
- 51:01Simulator integration with coordinator
- 55:24Q&A
related talks
29mBridging AI & Robotics: Exploring the LeRobot Library
20mProject Bonsai: How Microsoft is Using Ray and RLlib Inside a Machine Teaching Service
24mContext Aware Software Stacks for Mobility: Composive.ai Overview
11mSimAI: Unifying Architecture Design and Performance Tuning for Large-Scale Large...
13mLeveraging C++ for Efficient Motion Planning: RRT Algorithm for Robotic Arms
19mBoosting confidence in tiny AI deployment in embedded control systems
28mSAIDA RL: An Open-source Reinforcement Learning Platform
57mWriting Safety Critical Automotive Software for High Perf AI Hardware