While at Gibco our Senior VP introduced an interesting concept of Sneakerware to me. Sneakerware involved avoiding automation to gain a significant benefit in the efficiency of a lab.
Our dilemma was with bar coding at the time. The thought of the day way to convert all of the 96-well and 6-well plate feeding and passaging in the lab to systems with barcodes. The idea was that the new Hamilton STAR systems we had could be fitted with bar code readers, and by using complex programming, all of the plates could be tracked. We even went as far as exploring complete LIMS solutions from Elsevier and IDBS.
Our team gathered data and arguments, assembled a number of PowerPoints and headed into the VP's office.
He was a good VP. Heard us out. Asked to see a plate with a only bar code on it. Looked at us and asked, "What's in this plate?"
With a long pause we told him we did not know as we could not read the bar code yet.
His point was made with a stack of 10 plates and a sharpie. A brief label on the top plate and a diagonal line down the side of the other 9 was all it took.
"It's been that way for years and it works." he remarked. Well said.
Instead of large complex methods we chunked up the processes on the liquid handler. Brought a max of 10 plates at a time to passage then returned them by foot to the incubator.
That mix of Automation and Sneakerware did a number of things for the lab, including:
- Made the Automated Liquid Handler more accessible to more scientists.
- Reduced the complexity of the code by over half.
- Allowed for manual manipulation of the plates when the Hamilton was being used by others.
Many things can be done with automation. Simply put, some things should not be done through automation.
What's better, Sneakerware or Automation?
Hard to say. I do know that in the case of Gibco, the mix of the two worked out great when we used the right labware and technique for the application.