The Merck Automation challenge
This is a condensed version of the original story by Scott Ziegler, Director of Scientific Automation at Merck. Portions were contributed by Dr. Margaret Bryans, Montgomery County Community College.
A big difference between biotech in the classroom and biotech in industry is the use of robots. No company could routinely screen thousands of samples if it relied on people pipetting by hand.
But getting started with automation can be daunting. Montgomery County Community College introduced one cohort of students to automation through a three-day mini-hackathon–the Merck Automation Challenge.
Fortunately, Montgomery County Community College (MCCC) has instructors who are always ready to try new things and a program advisor, Scott Ziegler, who wants to see students learn automation-related skills.
The result?
The Merck Automation Challenge
Scott Ziegler, Merck’s Director of Scientific Automation, designed the challenge and recruited several Merck scientists to help.
Eight students participated in the event: four from biotechnology, two from engineering, and two from computer science. Instructors Dr. Maggie Bryans, Biotechnology, and Dr. Chengyan Wang, Engineering, were on hand to help.
Day 1 began at 9 am Saturday morning the weekend before Thanksgiving. Students practiced key techniques. In a 3-hour morning session, they pipetted 10 different volumes and determined the accuracy. They also performed an ELISA (enzyme-linked immunoassay) with four different samples and determined whether there was a response to a vaccine.
In a 3-hour afternoon session, students learned how to operate and program the Staubli 6 axis robotic arm. The Staubli representative taught them about safety, movement in 3D space, robot teaching, and the VAL 3 programming language.
In the next 3-hour session, they learned how to write code in Python and use Jupyter notebooks. On top of these skills, they learned about bar code scanners, bar codes, and RS-232 communications protocol. They also wrote code to control a relay over RS-232.
The challenge: After a 12-hour day, at 8:30 pm Saturday night, it was finally time for students to learn what the challenge was going to be.
They were told their task was to prove or disprove the hypothesis that automation is a wasted effort. To do this, they would be pipetting 10 different volumes into 100 vials. They would do 50 by hand and 50 using automation. For the automated process, they would use the Staubli robotic arm, a balance to weigh the vials, a bar code scanner to read the identifiers, and a Mantis robotic liquid handling device. At the end, they had to compare the pipetting accuracy for each person with each other and with the automated samples.
On day 2, 8 am Monday morning, the staff found some students already waiting and ready to go. During the day, students developed plans on whiteboards, presented them, and sent back to the board with new requirements and some limitations. The coaches checked in a few times during the day but did not tell them how do things. The last student left at 9:45 pm.
Day 3. Again, some students arrived an hour before the 9 am start time. They worked through the day with some monitoring by coaches and the belief that they had until 8 pm to complete their work. At 3 pm, they were told a vice president needed to see their work at 6 pm. The idea was to add a bit of “real world” pressure.
At 6 pm, they presented their work to a Merck vice president and discussed their findings. The day wrapped up around 9 pm. After 38 hours, students had learned the importance of cross-disciplinary work, and how computation, engineering, and biotech fit together. They also learned how to work as a team and experienced the pressure of timelines.
References:
1. Read the complete article by Scott Ziegler at the MCCC website.
2. See Staubli's robot in action.