Minnesota Science
Vol. 44, No. 2
Laboratory Robot Does Drudge Work With No Complaints
by Larry Etkin
[INTRODUCTION: Rainmakers and Robots: All to the Aid of Water Quality Research
In the public arena, scientists now debate the potential for, superconductivity and the possibility of room temperature fusion. Such debates focus attention on the process of research. It's a mysterious activity to many. We hear about the breakthroughs and the insights, but hove does science actually get done? Here's a look inside some experiment station scientists' labs, where researchers seek sometimes novel solutions to practical problems.]
==========
William Koskinen is a robotics innovator, but he's not a mechanical or electronics engineer. He's a soil scientist at the University of Minnesota. Recognizing the value of laboratory scale robots, he has set one up to expand the capability of his lab to keep pace with increasing demand for pesticide analyses of water and soils.
"We're studying the movement of pesticides through soil to groundwater," Koskinen says. "To be able to determine that, we have to do a lot of pesticide analyses. We have to extract the pesticide out of soil and water, and this is very expensive and time?consuming using manual labor," he says.
Koskinen's laboratory robot and one worker can prepare more than twice as many samples as three or four student workers employed full?time. It processes up to 200 samples a week and will pay for itself in about three years from the savings in labor, he says. Those savings are significant. The system cost $70,000.
"But even more important is that while the robot does everything in the preparation for pesticide extraction, and works all night, we've dramatically increased the precision of our work," Koskinen says. Sample variation of 15 percent with human labor dropped to 2 percent with the robot.
Development of the robot was financed by the Minnesota Agricultural Experiment Station and the USDA Agricultural Research Service. Koskinen says his system is relatively unique now, but won't be for long. "I think it's one of the first laboratory robotics systems in a university doing this kind of work. But it's the method of the future for doing a lot of this. With the number of samples to be analyzed continually increasing, you just can't afford to staff to do it anymore."
The robot itself is far more similar to heavy duty assembly line robots than to the humanoid robots of science fiction. It looks like a big table with a central arm and a lot of different stations set up around it. The arm picks up a tube with a soil and water sample and mixes it. It puts it into a centrifuge, pulls it back out, pulls out the extracted solution of pesticide, cleans it up, and runs it through a number of different steps. "The end product is a small amount of solution in a vial that's automatically capped and ready to go for analysis," says Koskinen.
"Our system has different hands to do different things. If it has to pick up a large tube, it goes and puts a big hand on the arm, and then if it has to pick up a small tube, it parks the big hand and picks up a small hand."
After a sample is prepared, a lab assistant carries it to a second lab where another automated system analyzes the sample. Essentially, it's a two?robot system with a human go?between and, of course, another human evaluating the final analyzed result.
PHOTO CUTLINE: LeEtta Jarvis teams up with a laboratory robot to help water
quality researchers analyze movement of pesticides through soils.