ROS 2, a successor of ROS, is a robotics framework suitable for programming of robotic, autonomous driving and other applications where distributed computing, large sensor data throughputs and multiple actuators are part of the system. ROS 2 is written in modern C++ 14 and is open source. In this talk we will for the first time ever present how such a framework has been certified according to ISO 26262, up to ASIL D, SEooC. We will present the following main challenges and how they were solved: How to retroactively derive the functional and safety requirements and trace them How to perform the safety analysis (FMEA) How to detect and eliminate the system memory allocation and blocking calls How to perform MC/DC code coverage in C++ 14 How we selected the target hardware and the RTOS platform How we performed the tool qualification and classification the non automotive tools We will conclude the talk with the analysis of the effort invested into this project and with the analysis of improvement of code quality of ROS 2.

http://CppCon.org — Discussion & Comments: https://www.reddit.com/r/cpp/ — Presentation Slides, PDFs, Source Code and other presenter materials are available at: https://github.com/CppCon/CppCon2019 — Using modern C++ is frowned upon by many working in safety critical and real time systems. We show our efforts to bring modern C++ into this new field on the example of a safe real time robotics framework for autonomous driving tightly integrated into ROS 2. This talk introduces the main challenges and how we solved them: * Avoiding memory allocations during exception handling. * Memory pools for standard containers and best practices to avoid memory fragmentation. * Additions to the thread support library to make it suitable for real time applications. — Andreas Pasternak Technical Lead, Apex.AI Working on algorithmic support C++ framework for autonomous driving vehicles. — Videos Filmed & Edited by Bash Films: http://www.BashFilms.com Work at Hudson River Trading (HRT): https://tinyurl.com/safxfctf

https://cppcon.org/ https://github.com/CppCon/CppCon2020/blob/main/Presentations/practical_memory_pool_based_allocators_for_modern_cpp/practical_memory_pool_based_allocators_for_modern_cpp__misha_shalem__cppcon_2020.pdf --- Runtime-deterministic memory allocations are a crucial aspect of any safety-critical real-time system. One of the simplest and widely adopted allocation mechanisms used in such systems is a memory pool with fixed block sizes. Unfortunately, the need to know the exact sizes of the memory blocks makes any practical usage of memory pools with standard C++ allocator-based approach rather problematic since users often “hide” real properties of allocations which are made under the hood. For example: STL’s node-based containers like 'std::map' as well as other standard mechanisms like 'std::promise' or 'std::allocate_shared'. Being a company which focuses on real-time safety-critical applications, we still see a significant value in keeping compatibility with the standard allocator model as well as in following common conventions which are familiar to every C++ developer. This talk presents an approach which uses a combination of a memory allocator implementation which instruments the code, and an external LLVM-based tool which extracts the instrumentation information and generates static memory pool definitions, allowing the allocator to switch from the heap to a memory pool without any further changes to the code. The presentation will walk through a simplest possible implementation of this approach. --- Misha Shalem C++ Architect, Apex.AI C++ developer with 16+ years of experience. Currently holds position of C++ Architect at Apex.AI, Palo Alto, CA --- Work at Hudson River Trading (HRT): https://tinyurl.com/safxfctf --- Streamed & Edited by Digital Medium Ltd - events.digital-medium.co.ukevents@digital-medium.co.uk *-----* Register Now For CppCon 2022: https://cppcon.org/registration/ *-----*

Autoware.Auto is an open-source autonomous driving stack built on ROS 2. This track aims at providing an overview of the capabilities to the ROS 2 community. Autonomous Valet Parking The Autoware Foundation has completed its first software demonstration with Autoware.Auto - Autonomous Valet Parking! But what is Autonomous Valet Parking and how did we achieve this milestone? How close are we to roaming the streets with a fully autonomous vehicle? What's next for the Foundation? Get the answers to the questions and more! Localization and State Estimation in Autoware.Auto This talk is about algorithms for localization and state estimation implemented in Autoware.Auto. It explains what localization and state estimation are, goes into details of the typical methods used to implement these concepts, as well as presents architecture decisions specific to the Autoware.Auto implementation. Object Detection and Controls in Autoware.Auto This talk gives a holistic overview over the 3D object detection stack available in Autoware.Auto. In addition it provides insight into the Model Predictive Controller and the Pure Pursuit Controller in Autoware.Auto. Behavioral and Motion Planning in Autoware.Auto This talk provides details about the architecture and algorithms for the planning module in Autoware.Auto. It explains how the behavior planner makes decisions from information provided by other modules in Autoware.Auto and how the motion planner plans the trajectory to a given goal. 200mph ROS - How Indy Autonomous Speeds Open Source University teams are exploring high speed autonomous driving via Indy Autonomous Challenge. We discuss technical challenges and how the open source community is rallying with autonomous driving stack of Autoware.Auto, ROS 2, OpenCV, Eclipse CycloneDDS with iceoryx and Zenoh V2X. #iac2021

Peter Baughman (Apex.AI), Michel Hidalgo (Ekumen), William Woodall (Open Robotics) In ROS2, writing automated tests is easier than it was in ROS because it is no longer necessary to coordinate a ROS master with the tests, however this is still a need for robust multi-process testing as well as testing of ROS2's new launch descriptions, which can contain arbitrarily complex operations. Launch Testing aims to provide rostest-like functionality for ROS2, as well as provide new functionality to allow for more robust integration testing. Video presented at ROSCon 2019. More details and slides are available at: https://roscon.ros.org/2019/.