From: J. R. Molloy (jr@shasta.com)
Date: Fri Apr 13 2001 - 12:19:21 MDT
Harvey Newstrom wrote,
> Wow, this looks interesting. When I first started reading it, I thought it
> was hype. I sometimes write software for clients, and I was thinking that
> these were just control messages over an internal network. But when I got
> to the line that explained that these weren't single-destination messages, I
> realized what was going on. Instead of sending messages from a source to a
> destination, I am guessing that they keep an simulated representation of the
> environment. Instead of directly instructing a piece to perform a function,
> they apparently adjust the representation of the different hormone levels.
> More than one object can effect these levels with source data, and more than
> one object can modify their own behavior based on these levels. This is
> almost like simulating a role-playing game character with various changing
> ability scores, and having the robot react as his current scores allow.
> This does sound like an interesting approach.
Sounds to me like you understand the operation (and implications) of Salemi's
work.
This technology extends to "self-configuring" as well as "self-scaling"
robotics:
http://www.trnmag.com/Stories/012401/Self-configuring_robot_mimics_lifeforms_0
12401.html
The modules connect via one of four docking ports located on four sides of
each module. The ports pair connectors with infrared communications systems
that guide the connection process and allow modules to exchange hormone-like
messages.
Connected modules communicate using these messages. In addition, modules from
different robots can also communicate this way via open ports that are 30
centimeters or closer. This allows robots to coordinate actions and exchange
modules.
----------------
The key to the robot's self-sufficient and autonomous behavior is the
hormone-like communication system, which allows modules to coordinate
movements and reconfigurations.
The modules can broadcast messages to the other modules that make up a robot.
These messages trigger the specific actions required for the robots to
assemble, move and change shape.
Like biological hormones, the messages last only a certain length of time,
trigger different actions in different receiving sites, and leave the
execution and coordination of a local action to the module performing the
action.
________________
Some folks imagine robots that exchange messages like organisms exchange
pheromones.
Huge teams of rather simple robots could build complex structures by
communicating instructions to each other the way ants do, via
olfactory/hormonal code.
Here's some hype about "reconfigurable robots"
The USC/ISI CONRO PROJECT
http://www.isi.edu/conro/
The base topology is simply connected, as in a snake, but the system can
reconfigure itself in order to grow a set of legs or other specialized
appendages. Each module will consist of a CPU, some memory, a battery, and a
micro-motor plus a variety of other sensors and functionality, including
vision and wireless connection and docking sensors. Major challenges include
packaging, power and cooling as well as the major issue of programming and
program control.
Like the man said:
"The only thing missing is most of the knowledge
we'll need to make such machines intelligent. Indeed,
as you might guess from all this, the focus of re-
search in Artificial Intelligence should be to find good
ways, as Vinge's fantasy suggests, to connect struc-
tures with functions through the use of symbols."
--Marvin Minsky, Afterword to True Names, 1984
Some experimenters have gone so far as to equate robot hormones to emotions:
Artificial Emotions as Emergent Phenomena
http://fox.nstn.ca/~aai/roman93.htm
Although some researchers claim that emotion is unique to mammals, this paper
describes a notion of artificial emotion as a phenomenon resulting from a
series of modifications to emergent behaviors generated by a behavior-based
artificial intelligence (AI) approach. Such modifications to behaviors are
caused by stimuli (including those from humans) which a robot receives from
its environment. The paper describes a series of experiments to generate and
test artificial emotion using Subsumption Architecture (SA) robot platforms
developed by the Massachusetts Institute of Technology (MIT). A "hormone
mechanism", which is part of the behavior definition language, was used to
generate artificial emotion. In addition, the Action Selection Dynamics (ASD)
paradigm proposed by Pattie Maes as a way to implement Computational
Reflection was also tried. The latter is expected to permit us to investigate
more profound ontological issues associated with artificial emotion as part of
the experiments in Computational Reflection.
Stay hungry,
--J. R.
Useless hypotheses:
consciousness, phlogiston, philosophy, vitalism, mind, free will, qualia,
analog computing, cultural relativism
Everything that can happen has already happened, not just once,
but an infinite number of times, and will continue to do so forever.
(Everything that can happen = more than anyone can imagine.)
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:06:57 MST