IEEE Robotics & Automation Magazine - June 2021 - 103

Other works have been presented to enable visual programming
of robots through state machines [14], [15]. One
example, which is comparable to our proposed framework, is
RMC Advanced Flow Control (RAFCON) [16]. This hardware-
and middleware-independent framework allows the
user to design complex robotics tasks through an intuitive
GUI. It uses a system of hierarchical state machines to determine
and coordinate the robot behavior.
RAFCON is a very general framework for programming
robots, ranging from simple indoor navigation to complete
space-exploration missions. To handle such complex sequences
of operations, the task description is high level (e.g.,
" explore " or " pick object " ) rather than low level (e.g., and uses a
specific inverse kinematic solver to move to a computed position).
Given the variety of tasks and robots that can be programmed,
the practical use of this tool becomes very complex
and may be dependent on previously developed examples. In
particular, the integration of low-level components may not be
straightforward.
Some industry-specific frameworks are also commercially
available and allow for rapid and intuitive design of automated
and/or autonomous tasks, such as Artiminds (https://
www.artiminds.com/), Mech-Viz (http://en.mech-mind.net/),
and drag&bot (https://www.dragandbot.com/). These systems
mainly support industrial robots and can be programmed
through a drag-and-drop interface and come with
a set of available operations (e.g., move or search).
However, these systems are not always able to accommodate
external components (e.g., from academia), as companies
usually offer extensions for autonomous, sensory-based
systems. We have, therefore, identified
the need for a framework that can
incorporate both low- and high-level
components in an intuitive way.
This is the aim of GRIP: a hardOptional
Manual
Integration
ware-agnostic, ROS-based software
framework that facilitates prototyping
and robot integration specifically
for grasping and manipulation tasks.
Both low- and high-level components
(kinematic libraries, controllers,
motion planners, sensors, and learning
methods) can be integrated without
imposing constraints on their
implementation (see Figure 1). This
greatly improves the reusability of
existing software components in the
industrial setting. Unlike the existing
solutions that mainly focus on easing
the programming part, GRIP guides
users from robot and software integration
to task execution thanks to a
reactive GUI.
The architecture of our framework
minimizes the effort required to interface
with existing software, even on
ROS Libraries
Can Configure
Methods Manager
Hardware Manager
Task Editor
Can Configure
Launch File
Can Be
Integrated to
Low-Level Components
Can Configure
MoveIt!
ROS
Optional Manual
Integration
Figure 1. An overview of the GRIP framework. The architecture allows users to operate
robots with integrated components (hardware and/or software) for grasping and
manipulation tasks. The task editor provides an intuitive interface for designing and
tuning the robot's behavior. Arrows indicate the different (mutually compatible) ways to
interface external components.
JUNE 2021 * IEEE ROBOTICS & AUTOMATION MAGAZINE *
103
complex systems (e.g., bimanual robots). We therefore believe
that GRIP is a useful tool that makes it possible to evaluate
grasping and manipulation software/hardware with minimal
overhead. The optimal combination of components for
a given task can then be integrated into an existing and
optimized pipeline.
Overview of GRIP
The main purpose of this framework is to provide an environment
in which robots can be operated with ROS-compatible
software components coming from academia as well as
external software components. Although GRIP provides a
wide range of integration and interfacing options, we made
sure to keep its use straightforward and as programming free
as possible, as follows:
●
●
robot interfacing, from material widely available online
software and sensor integration, regardless of their implementation
details
●
●
definitions of variables to be used during the robot
execution
task design and execution with integrated components
through a visual drag-and-drop interface.
Typical use cases implemented using this procedure are
described in the " Integration " section, and videos are
available online (https://sr-grip.readthedocs.io/en/latest/).
Integration
This section explains how GRIP is able to run a wide range
of hardware and software to make good use of existing
components. This entails two challenges: 1) how to avoid the
High-Level Components
Can Be
Integrated to
Configures
Generates
ROS Messages
Handled by
Message Managers
Setup Configuration
GRIP
http://www.artiminds.com/ http://www.artiminds.com/ http://en.mech-mind.net/ https://www.dragandbot.com/ https://sr-grip.readthedocs.io/en/latest/
IEEE Robotics & Automation Magazine - June 2021

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - June 2021
