IUBio

Robotics

Douglas Douglas
Thu Jan 30 10:17:49 EST 1997


Hi everyone,
   We have a lot of robots here, from a number of vendors.  My background
is in Mechanical Engineering, M.S. in Controls and Robotics.  The robots
we currently use are:  Packard MultiProbe 104 and 204 (not disposable
tip), Hamilton MicroLab 2200, PE ABD Catalyst, PBA Flexys, Robbins Hydra
96 (not really a robot), Sagian MultiPette (96 channel pipettor - 6 deck
positions), Biomek 2000, CRS T265 (smaller A255 arm on a track).  

   The robots that get used the most are the Packards.  They are easy to
program and can be customized with a good engineering group (see my
abstract for my poster at LabAutomation '97).  They have been very
reliable (note we do not use disposable tips so I don't know about that
configuration). The Biomek 2000 is very good with procedures that need a
variety of tools, but has the major flaw that it can not ask the user for
input during a run, or dynamically reallocate the procedure for the number
of samples (like the Hamilton, Tecan or Packards can).  If it is the same
procedure every time, they are excellent.  Further, they have the best
disposable tip handling and options. The Hamiltons are limited in
programming capability and were clearly designed for lower throughput than
we use them for.  We have needed extensive preventative maintenance for
them.  Admittedly, they are being used 10-12 hours solid every day.  The
Robbins has been a reliable work horse; however, replacing plunger tips
can be a pain.  The PBA Flexys (a picking and gridding robot) has been
excellent for gridding and more of a problem for picking.  I hear that a
new picking tool is now available and is more reliable.  The MultiPette
has a very easy to deal with programming system.  It is reliable down to
15l (with the P20 disposable tips) and is already set up to handle 384
well plates.  Once again it is not able to dyanmically ask the user for
volumes or plates, however the running files are all text.  This allows a
good programmer to write a user interface that edits these files and then
runs the system via the Windows DDE links.  I have extensive experience
with industrial robot arms (notably Panasonic and Accusembler with some
GMFanuc experience).  I find the CRS reliable, easy to program (for an
industrial arm!) and the company willing to respond to concerns of a
customer who will order only a few units.  The software also allows for
easy communication with other Windows based systems via DDE links. 
Lastly, the Catalysts are also work horses.  They run off-the-shelf and
all the time.  The only warning is to clean the PCR block twice a month
and replace syringes once a week if you are running them all the time
(like we do).  

   Now for the systems we don't own.  We had an initial negative interlude
with the Tecan Genesis just after it was introduced 15 months ago.  I have
not tried them since so I don't know what they are like now.  We do like
the RSP/MiniPrep systems and are ordering CAVRO syringe drivers for a
number of custom applications.  The Gilson does what is promises (septum
piercing, etc.) but was too slow for our needs.  We have not tried the
Q-bot, CyberLab or PSS systems.

   In terms of arms (a subject I know well) we find the ORCA a little
under-powered and small for some of our needs, but the easiest to use for
labs with limited engineering personnel.  The MItsubishi's which are being
advertised so heavily by Rixan are inexpensive, but are kind of a step
child product.  Their payloads and accuracies are in between the ORCA and
the CRS arms.  Note that they were developed for IN-HOUSE production in
Japan because big Japanese Karetsu (allied conglomerates) are in general
not allowed to buy a competitors equipment!  Even the Rixan salesman said
they were developed to produce industrial electronics that Mitsubishi
sells for $9million/month.  Some of the Mitsubishi sales force convinced
corporate to sell them to others, but it is only a $9million/year business
for a $40billion/year company.  Not a lot of support.  Another choice in
this category is the Panasonic (Matsushita) PanaRobo KS-V20, which I have
heard rumors is now controlled by an Adept one controller.  If the Adept
controller is indeed being used, then the control side is handled by an
American automation company with a large installed base.  
  
   For more information on robotics for the lab, please contact the
Association for Laboratory Automation and subscribe for the Laboratory
Automation News magazine at P.O. Box 572, Charlottesville, VA, 22908.  You
can also fax or email the president Dr. Robin A. Felder at (804) 924-5718
or rfelder at virginia.edu.  The big ALA meeting is LabAutomation, which for
1997 was held in San Diego last week!

Sincerely,
Douglas Gurevitch

-- 
**************************************************************************
The opinions expressed here do not represent the opinions of SEQUANA 
THERAPEUTICS.
Douglas Gurevitch  -   douglas at sequana.com
**************************************************************************



More information about the Autoseq mailing list

Send comments to us at biosci-help [At] net.bio.net