Subject:
|
Visit to MDRobotics/SPAR
|
Newsgroups:
|
lugnet.org.ca.rtltoronto
|
Date:
|
Fri, 27 Sep 2002 03:39:29 GMT
|
Viewed:
|
391 times
|
| |
|
|
Hey folks,
Last night I visited MacDonald Dettwiler Robotics (MDRobotics), formerly
known as SPAR Aerospace, the Canadian company that brought you the Canadarm
Remote Manipulator System. The Toronto chapter of INCOSE, the International
Council on Systems Engineering, had their bimonthly meeting there and MDR
gave a talk on developing new arms for Mars exploration, as well as gave us
a tour of some of their facilities. I ran into John Guerquin at the
University and invited him to come along.
SPAR was bought out two years ago by MacDonald Dettwiler, an American
company. They've always been, like IMAX and even Nortel for that matter,
the prototypical great Canadian engineering firm--if you ask any child of an
example of Canadian engineering-they'll immediately say the Canadarm. Their
first product was the Canadarm, or the Remote Manipulator System. That's
the robotic arm inside the US Space Shuttle. They've made five of them, one
per shuttle, and regularily get repaired and serviced at MDR in Brampton.
While we were there, one was being serviced in many pieces after an accident
where someone at NASA apparently dropped something on it. D'oh!
Their second product is the SSRMS, or Canadarm2 (I remember Discovery had
this naming contest for the damned thing AND THE BEST NAME THEY CAME UP WITH
WAS CANADARM2? Jesus.) The SSRMS is the multijointed arm for the space
station that was delivered and installed by Chris Hadfield on STS-100.
It's different from the RMS in that it's much larger, heavier and can carry
larger objects. It's also better because it takes advantage of new
technology which allows for better user interface control and autonomous
operation. The SSRMS can inchworm to different locations on the station,
and is attached to a train car which slides along the main truss.
One of the managers there said that the SSRMS has logged more flight hours
and usage in a year and half of being up there than all five RMS arms have
in 20 years of shuttle operations. The SSRMS apparently get an accuracy of
1.5mm at its full length of ~60-70 feet long, while in space with
temperature ranges, radiation, micrometeor debris etc. Imagine the accuracy
we get with ProjectX! The guy hammered this in by pointing his laser
pointer at the back wall of the room, and noting he can't even keep the
pointer steady within inches at 30 feet.
The SSRMS incorporates MDR's third product, the Special Purpose Dexterous
Manipulator (SPDM), or "Canada Hand". This is a smaller machine with two
little fingers which is used to grab stuff and manipulate it. They showed
us the giant test rig which was used to test the SPDM-apparently you can't
test the range of motion entirely because of gravity, so pulleys and rigs
are used to simulate zero gravity for the hand. The SPDM is undergoing
testing right now-we didn't get to see the actual unit itself, but will be
installed on the ISS in a few years.
Visiting MDR itself was a real treat. It was like going to the Science
Center, but the real thing. The place looks just like Science Center
too-they have a model of the Space Shuttle in the lobby, and a model of the
ISS truss segment with the SSRMS on it. On the SSRMS model is, get this, a
Star Wars stormtrooper hanging out on the middle joint. :) They also had a
1/3" model of the grappler on the end of the RMS-a wire iris which grabs
satellite mounting points. Inside we got to see the high bay which you have
to wear the little booties and hair nets, as well as experiment areas.
Later in the evening they showed a video of STS-100 where they installed the
SSRMS using the old RMS arm. In the video Chris Hadfield talks about a
"unique configuration"-the folks from MDR told us that was their name for
aligning the two arms in space so that they could take a photo of them with
the Canada logos facing together!
The other major demo was of a new project-a robot arm for future NASA Mars
missions. This arm is only about six feet tall and has 6 degrees of
freedom. It's meant to sit on Mars lander or a rover and grab experiment
packages and scoop up samples for analysis. We got to look at it up close
and watch it run through operation-it grabbed a small box about the size of
a milk carton and moved it across, then grabbed a scooper and scooped up
some sand in a sandbox to simulate digging on another planet.
The machine is made of lightweight and strong materials-parts are made of
carbon fibre and aluminum beryllium, so that the entire package-both
electronics, motors, structural elements, etc is only 9 kg. One of the
things they showed was a integrated sensor-a potentiometer built into a
ring, which was fabricated into an alumnimum part. The part is
structural-it fits between bones in the arm-yet to save weight, it also
gives a secondary sensor readback to the computer.
What's really nifty about the small arm is that it is heavily distributed in
its control. The arm is driven from an electronics package which runs Wind
River Systems' VXWorks operating system on a radhardended flight ready CPU
board. This part sits on the lander. The ground segment back at NASA is a
normal PC which controls and sends commands to the electronics flight
package. But what's different in this is that most of the operations are
closed loop and autonomous. The ground only sends up a small command that
says "Go here" and it's up to the arm and it's flight electronics and
software to figure out how to do it. For example, if an object is not
exactly where it's supposed to be (Martians moved it?), you don't have to
wait the 30 minute round trip time lag to tell the robot to fumble around
for it. The arm autonomously tries to correct for the error and hunts
around in a predefined area for it.
A good part of the evening was a talk about the systems engineering
methodologies and processes they used to develop the new arm. They took a
small team (10 engineers) and a small budget ($2.5M-well, small is relative)
and did the entire project in 15 months, on budget and on schedule! Now
that's project management. What's interesting to note is the constraints
and objects they had were pretty stringent-power usage, space, weight etc so
that it would fit onto a Mars lander. Next time, I don't want anyone to
complain about rtlToronto event rules!
Anyways, it was a pretty cool visit-well worth it. I might join up with
INCOSE just to keep going to these things.
Calum
|
|
Message has 1 Reply:
 | | RE: Visit to MDRobotics/SPAR
|
| (...) Overkill on the OS, but then again it's off the shelf so you don't need to spend 1 or 2 man years designing, coding, and testing.... (...) Yes!!! Small teams of dedicated smart people can make all the difference. Cheers, Ralph (22 years ago, 27-Sep-02, to lugnet.org.ca.rtltoronto)
|
3 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
