fpgaDDS: An Intra-FPGA Data Distribution Service for ROS 2 Robotics Applications

Christian Lienen,Sorel Horst Middeke,Marco Platzner

2023-03-01

Abstract:Modern computing platforms for robotics applications comprise a set of heterogeneous elements, e.g., multi-core CPUs, embedded GPUs, and FPGAs. FPGAs are reprogrammable hardware devices that allow for fast and energy-efficient computation of many relevant tasks in robotics. ROS is the de-facto programming standard for robotics and decomposes an application into a set of communicating nodes. ReconROS is a previous approach that can map complete ROS nodes into hardware for acceleration. Since ReconROS relies on standard ROS communication layers, exchanging data between hardware-mapped nodes can lead to a performance bottleneck. This paper presents fpgaDDS, a lean data distribution service for hardware-mapped ROS 2 nodes. fpgaDDS relies on a customized and statically generated streaming-based communication architecture. We detail this communication architecture with its components and outline its benefits. We evaluate fpgaDDS on a test example and a larger autonomous vehicle case study. Compared to a ROS 2 application in software, we achieve speedups of up to 13.34 and reduce jitter by two orders of magnitude.

Robotics

What problem does this paper attempt to address?

The paper mainly addresses the computational demands and performance bottlenecks in modern robotic applications, particularly focusing on the data transfer efficiency when using the ROS (Robot Operating System) framework in combination with FPGA (Field-Programmable Gate Array) hardware acceleration. Specifically, researchers have found that the traditional ROS communication mechanism encounters bottlenecks when handling data exchange between FPGA hardware acceleration nodes. This is primarily because all data needs to be transmitted through shared main memory, leading to unnecessary delays and performance degradation. To solve this issue, the paper proposes fpgaDDS (Field-Programmable Gate Array Data Distribution Service), a lightweight data distribution service specifically designed for hardware-accelerated ROS 2 nodes. fpgaDDS is based on a customized streaming communication architecture and integrates with the ReconROS architecture and programming model, aiming to improve data transfer efficiency between FPGA hardware acceleration nodes. This service can significantly reduce the latency of inter-node communication and decrease jitter during system operation, which is crucial for achieving high-performance, low-latency robotic systems. In summary, the main goal of the paper is to optimize the data transfer efficiency between FPGA hardware acceleration nodes in ROS 2 applications by proposing fpgaDDS, thereby enhancing the overall system performance. Experimental results show that compared to software implementations, using fpgaDDS can significantly increase execution speed and greatly reduce system jitter.

fpgaDDS: An Intra-FPGA Data Distribution Service for ROS 2 Robotics Applications

Mapping and Optimizing Communication in ROS 2-based Applications on Configurable System-on-Chip Platforms

ReconROS Executor: Event-Driven Programming of FPGA-accelerated ROS 2 Applications

Fuzzy DR Algorithm for Data Distribution Management

Performance Evaluation of ROS2-DDS middleware implementations facilitating Cooperative Driving in Autonomous Vehicle

RobotCore: An Open Architecture for Hardware Acceleration in ROS 2

ROS 2 on a Chip, Achieving Brain-Like Speeds and Efficiency in Robotic Networking

FogROS G: Enabling Secure, Connected and Mobile Fog Robotics with Global Addressability

A High-Performance FPGA-Based RoCE v2 RDMA Packet Parser and Generator

A Hardware-Software Co-design based Approach for Development of a Distributed DAQ System using FPGA

Comparison of Middlewares in Edge-to-Edge and Edge-to-Cloud Communication for Distributed ROS2 Systems

FogROS: An Adaptive Framework for Automating Fog Robotics Deployment

Security and Performance Considerations in ROS 2: A Balancing Act

A Survey of FPGA-Based Robotic Computing

FABLE-DTS: Hardware-Software Co-Design of a Fast and Stable Data Transmission System for FPGAs

Elastic-DF: Scaling Performance of DNN Inference in FPGA Clouds through Automatic Partitioning

AutonomROS: A ReconROS-based Autonomous Driving Unit

Skew-Oblivious Data Routing for Data-Intensive Applications on FPGAs with HLS

Optimizing Neural Network Inference in Edge Robotics by Harnessing FPGA Hardware Acceleration

Multi-Stream Scheduling of Inference Pipelines on Edge Devices - a DRL Approach

Architecture exploration of intelligent robot system using ROS-compliant FPGA component