Show EOL distros: 

Package Summary

Package which implements capabilities, including code to parse capability interface specs, to parse capability provider specs, and implement the capability server.

Package Summary

Package which implements capabilities, including code to parse capability interface specs, to parse capability provider specs, and implement the capability server.

Package Summary

Package which implements capabilities, including code to parse capability interface specs, to parse capability provider specs, and implement the capability server.

Package Summary

Package which implements capabilities, including code to parse capability interface specs, to parse capability provider specs, and implement the capability server.

Package Summary

Package which implements capabilities, including code to parse capability interface specs, to parse capability provider specs, and implement the capability server.

Package Summary

Package which implements capabilities, including code to parse capability interface specs, to parse capability provider specs, and implement the capability server.

Package Summary

Package which implements capabilities, including code to parse capability interface specs, to parse capability provider specs, and implement the capability server.

Description

The notion of Capabilities in ROS is that there exists a higher level interface for most "capabilities" that robots possess. An example of this is the organic evolution of interfaces for things like the Navigation Stack and the tf frame names. These sane defaults were used on robots like the PR2 and have become defacto standards over time, and some have even been solidified in REP's. But there has never been a system in place to define these higher level interfaces (which consist of other ROS concepts such as topics, services, and parameters) and to manage their runtime in a way that makes interacting with an unknown or abstracted robot easier.

One of the main goals of the capabilities is to make writing robot agnostic "apps" easier. If a robot can declare that it implements the standardized interface for the navigation capability, then an app can rely on the ROS interface used to utilize that capability. Making it easier for app developers to work independently of the robot developers and making it easier to integrate onto new robots. This is analogous to the use of Messages in topics and services, and that has proven to be very effective at easing integration of disparate parts of the system.

Documentation

Tutorials

There are tutorials located here: capabilities/Tutorials

API Documentation

There is fairly extensive API documentation for this package in the CodeAPI link on the right side of this page.

The provided (Python) API can be used to discover local packages which define one or more of the spec files defined by this package, load them, and interact with them. It also provides an API to discover what capabilities are provided by a capability_server running on a remote machine, which is useful for tools like the rqt_capabilities plugin.

capability_server

This package provides the capability_server ROS node which loads all capability spec files from packages on the ROS_PACKAGE_PATH and provides a ROS interface for querying and running those capabilities.

The capability_server is documented in the CodeAPI (linked to on the right).

Design Documents

There is no organized design documentation, but there are a few documents used during the design process that might be useful for anyone wanting to understand the existing work:

There was also an article in IEEE Spectrum (ROS Topics):

Wiki: capabilities (last edited 2018-01-13 15:01:24 by NickLamprianidis)