Advances in networking,
communication and computation technologies present several exciting
possibilities for distributed robotic systems. In the past years a number of
research efforts have been carried out to provide with Internet based robotics.
Most of these involved teleoperated arm robots [R1
such as the Mercury project [R2
Bradford Telescope [R4
Australia´s Telerobot [R5
the Internet based remote control for the Pathfinder [R7
and many others. A number of mobile (and sometimes autonomous) robots were also
adapted for web control, such as Xavier [R8
interactive museum tour guide robots (Rhino [R9
and Minerva [R10
as well as others.
These projects have highlighted
the potential of Internet when applied to robotics as well as some of its
limitations. Among the shortcomings experienced were (1) limited bandwidth
(original Internet), (2) restrictions in wireless transmission, usually radio,
over the area of operation of the robot and (3) restricted tasks to accommodate
these shortcomings. With the introduction of Internet2
and ubiquitous wireless
communication, that will soon handle real time video, it becomes now more
realistic to develop an Internet based mobile robot laboratory that will enable
more sophisticated mobile robot architectures accessible to a more extensive
audience. In particular, advances in IP-based
communication for handheld
devices equipped with wireless interfaces is creating new challenges for mobile
middleware frameworks, while opening new possibilities in areas such as
distributed robotic system (DRS), where intelligent autonomous mobile robots can
communicate and collaborate in order to perform complex tasks. However,
Distributed intelligent and autonomous (remote) robots not only demand
ubiquitous access to information (anywhere, anyplace and anytime), but also a
high degree of flexibility and adaptability not present on current communication
frameworks, in order to deal with changes in the computing and communication
environment. For example, logical mobility, or the movement of code and data, as
well as physical mobility may affect network connection (disconnection, reduced
connectivity) while DRS is running. In this scenario, the robot needs to be able
- Detect and adapt itself to the change of location (location
- Adapt to changing network and environment conditions (transient failures,
disconnection, or reduced connectivity) due to power consumption, available
spectrum and mobility.
- Integrate power aware mechanisms in order to determine the
parameters required to achieve cost-effective system performance and
self-adaptation based in robot's constraints and configuration, as well as
communication service availability.
In such a way, the MIRO project,
Adaptive Middleware for Mobile Internet Robot Laboratory involves four main
research areas and development thrusts:
These areas are linked together
by a distributed architecture as shown in figure 1.
MIRO Distributed Architecture.
The model and simulation repository as well as the distributed
NSL/ASL system, distributed in a set of workstations connected to the
internet2 infrastructure, provide the robot's intelligence; while the robot
itself is located in a wireless environment. The middleware framework
facilitates communication between the networked workstations and the mobile
robots, hiding mobility complexities; while extending robot's mobility and
The specific objectives of
this joint research project are fourfold:
- To provide an understanding and means by which Internet2 can be
efficiently integrated into a public wireless network of single and multiple
autonomous mobile robots capable of handling real-time video.
- To provide an adaptive communication environment that will make it
transparent to the application and robot what the actual network
characteristics are and how to deal with the inherent restrictions.
- To provide a Internet2/wireless "grid" that can be effectively applied to
biologically inspired autonomous mobile robots linked to distributed
computational resources in the Web.
- To provide with a distributed virtual laboratory enabling real time
interaction with autonomous mobile robotic systems to users anywhere in the
In order to achieve these goals, the following issues will be
Environment: At the application level, the impact of mobility in
communication intensive tasks, such as video processing in wireless devices,
should be hidden by enforcing communication policies that allow the robot to
adapt in response to an environmental or system configuration change.
Although, how, when and what to adapt is often stated as
application dependent, this has an impact on the complexity of the
architecture and should be thought carefully. At the network level, the
communication restrictions and the selected communication protocol will
determine the attainable visual processing complexity (colors, gray scale,
horizontal/vertical rectangles, squares, etc.). Under these circumstances,
provisions should be taken in order to switch communication protocols and
policies according to the network fluctuations.
and location awareness: In order to best distribute processing and control
of mobile robots accessing Internet2 software facilities through a public
wireless network, a robot need to be able to track its position and the
position of other robots in open fields (GPS, "shadow" robots, etc.) and
monitor their power consumption. However, accuracy and frequency of the
operations should be studied carefully to determine when they are