Another "World's First" for a Robot?
Posted 13 Jun 2003 at 14:16 UTC by steve 
Taking a page out of the marketing plan of robot vacuum cleaner
companies, Smart Robots, maker of the SR4 and SR5 robots, claims
"Tuesday June 10, 2003 may well go down in history as the date when
autonomous mobile robots first began exchanging email with each other
over the Internet." While the SR4/5 robots do look like cool
robots, their marketing folks may have gotten a little carried away.
What about all the other robots that have been sending and receiving
email for years? Like the Toshiba ApriAlpha, Evolution ER1, Sony
SDR-4X, AIBO, and most Linux-based robots. Anyone have a guess at what robot
was really the first to send and receive email?
Wouldn't it be great to create a robotic history timeline?
Date Event
x First Robot Vacuum created
x+1 First Emailing Robot created
x+2 First Robot Vacuum created
x+3 First Emailing Robot created
etc
I recall, probably around 1994-1996, a robot at a university that
wandered a hallway. There was a web site that you could access to give
the robot a task - you got to pick which office it went into, and the
action that it would perform. One of the actions was to either tell a
joke or tell a limerick (I've forgotten. I think it was a limerick,
though).
When the action was completed, the robot sent you an e-mail that said
that it had completed its task. I think I got my e-mail a day or two
after I'd sent the request.
xavier, posted 18 Jun 2003 at 13:04 UTC by jiggersplat »
(Journeyer)
the robot you are thinking of is xavier from carnegie mellon.
http://www-2.cs.cmu.edu/~Xavier/
We certainly did not intend to mislead with the article. We have
assumed that any number of robots have sent out open-text e-mail in
the past. However, I do believe that this is the first occasion of
robots actually exchanging e-mail (which is what our article says) and
being able to do so in a way that the robots can parse the contents of
the e-mail and act on that contents (using XML). If you find me to be
incorrect in this, I will certainly acknowledge that misperception.
We also used the word ``may'' and hope that we will, in fact, learn if
anyone else is doing this right now --- or has, in the past. Who has
really done it all first is not terribly important.
Our overall interest is to encourage others to investigate XML and web
services as the very potent robot-communications medium we believe it
is.
See robot to robot (http://www.smartrobots.com/robots_talking.jsp) and
send a robot an e-mail (http://www.smartrobots.com/email_a_robot.jsp)
on our site. We will continue to develop these techniques and make
API's, an XML-reader and examples available on our site as we go
along. We encourage your participation and efforts with us in the
open robot network we are developing (see
http://www.smartrobots.com/network.jsp).
Althought robots have collaborated and communicated
with each other for ages, it is nice that a robot can send and
receive
XML emails regardless if the sender is a human or a robot. The thing
is, is that it still seems a little misleading is that:
- It's not the first time that robots have communicated with other
robots? (this has gone on forever as shown by the link above)
- It's not the first time two devices have communicated by email or
XML (this
has happened for ages on the internet)
- It's not really communicating by human readable email (since you
have to format the emails into an XML format)
- It may be that it's the first time two robots talk by XML email
(but this is the most inefficient method for two robots to communicate
with each other so if it hasn't been done it's because it's that most
projects abandon it or choose rather to use
RPC or remote object communication or some other faster or more
efficient machine level form of
communication across network protocols and only use human readable
form for when they want to talk to humans).
It is handy that you can create an XML email and send it to a robot.
It's also handy that a robot could create a similar XML email and send
it back or even to another robot. It's
sort of weird that anyone would try to make someone think this is a
novel enough "first" (if even it is a first) for the world of robotics
that we should all jump up and down, take note, and put this
particular idea on some kind of pedestal.
It is cool, shrug, but not THAT cool. I'll agree with steve when
saying the article seems to get a little carried away.