H-ROS, the Hardware Robot Operating System, is a collection of hardware and software specifications and implementations that allows creating modular and distributed robot parts. H-ROS heavily relies on the use of ROS 2.0 as robotic development framework. This gives access to a collection of tools modern roboticists require for their daily work.
In addition, we have created an infrastructure that enables to fully exploit the potential of distributed systems and added the state-of-the-art electronic and communication techniques.
These are the technical key-benefits the use of H-ROS brings to robot parts:
Plug & Play: Create robot components that do not require any configuration. Just add or remove hardware parts easily and discover them in your robot network. H-ROS relies on Ethernet communication network to connect robot components.
Interoperable: Using the HRIM information model, a common interface is defined for each type of robot part. This allows using different manufacturers modules seamlessly, not requiring any additional coding to support them.
Extensible: Extend your robot adding new modules dynamically to adapt it to the use case. The robot parts are easily attached thanks to the use of Ethernet connectors.
Reconfigurable: Once the robot detects these new parts, it can adapt its behavior to new tasks. This allows the adaptation of the robot to which robot components it contains, dynamically.
Real-Time: H-ROS uses a Real Time Linux Operating System that allows determinism in the software execution and in the communication exchange. This is achieved using several Time Sensitive Network (TSN) related standards and the use of sub-microsecond synchronized clocks using Precision Time Protocol (PTP).
Security: H-ROS is built with security in mind. It uses encryption and authentication methods when running algorithms and exchanging data. Penetration tests are being performed to identify weaknesses and correct them.
Introspection: Obtain robot part's information and state thanks to the access H-ROS gives. This allows having first hand data at any time using ROS topics.
Automatic updates thanks to OTA: With Over-The-Air updates, the robot part software is updated easily.
Networking control: Shape robot's inner traffic reserving bandwidth for high priority data and work under loaded networks.
This and much more is possible thanks to the H-ROS System on Module (SoM), an electronic device that embeddeds all the required electronics to run modern robotic software and that turns your robot parts into completely modular.
Thanks to the H-ROS SoM, the augmented advantages of the module are:
ROS 2.0 Hardware: A purely distributed architecture built with ROS 2.0 that makes the module a first class-participant of the ROS 2.0 ecosystem.
Hardware level life-cycle: ROS 2.0 life-cycle extension to hardware which allows to influence power conditions for an increased performance or adaptive behavior.
Security: An encrypted computing and communication environment. A hacker-tested robot module, secured through continuous security audits.
Interoperability: A common interface, the Hardware Robot Information Model (HRIM), enables communication among different vendors regardless of the manufacturer. It also enables robot modules to be added or removed from the network without interfering with the runtime operation of any other device on said network without impacting any data flows in which they are not directly engaged.
Diagnostics and telemetry: First hand data about robot part states through ROS topics, such as power consumption, processor load and much, much more.
Automatic updates, over the air (OTA): Over-The-Air updates for robot parts. The SoM keeps robots and robot modules updated, seamlessly.
Full synchronization: Distributed sub-microsecond clock synchronization accuracy. ROS 2.0 latencies get optimized.
Real-Time Operating System: Deterministic operating system responses powered by a hybrid architecture featuring the most popular OS in the robotics domain: Linux.
Resilient networking: Applications that operate predictably in the presence of network congestion. Even with traffic bursts above 90% of the channel capacity, our solution delivers.
Traffic shaping: A mechanism that allows reserving bandwidth for high-priority traffic while, at the same time, ensuring best effort traffic will continue flowing.
Policing: Robot modules that meet their specifications by applying individual policy rules. This implies that even if your robot component malfunctions, we ensure that it does not compromise the rest of the robot network.
Bandwidth allocation: The option to dynamically estimate the available bandwidth in the network and to reserve an additional portion, if possible. This empowers roboticists with the status of the robot network.
Redundancy: A variety of different network architectures that enhance the redundancy of a robot network (daisy-chain, line, ring, tree, etc.), gaining reliability.
Low cycle times: Scalable and user-selectable cyclic update rates that can meet or exceed legacy Industrial Ethernet solutions.
Scalability: The possibility to grow from individual modules to large robots composed of hundreds of modules in multi-network setups.
Converged networks: Coexistence with best effort traffic and support to multiple industrial protocols.