Tecan has a fairly new feature called PMP, to my knowledge, it is similar to Hamilton’s TADM, however, less used.
I’m looking into intergrading it into our new fluent system, and would like to hear if anyone has any experience using it, has any documentation on it, or any other insightful comment about it?
Tecan is actively recruiting early adopters, but a paid upgrade to both hardware and software is required. Your local sales representative should be able to connect you with someone on the software development team who can give you a detailed overview.
I’ve used it on the EVO’s but I remember noticing that the hardware on the Fluent had the PMP option on the airFCA board for a few years now.
FluentControl 3.3 released in December 2022 brought new PMP to the Fluent.
Yeah there was already the capability on the board for the Fluent FCA arm but it was just missing the hardware. When I did the Fluent hardware training back in the day (2018?) it was pointed out.
Also check this out: Tecan’s Pressure Monitored Pipetting supports Swiss BioQuant’s dedication to excellence
I’m definitely curious about it
Did you end up purchasing & integrating it into your Fluent system?
We have three systems with the PMP license. So far PMP is activated to record data on two critical transfers, 2µl transfer of sample and 18µl of PCR reagent.
At the moment we are manually checking if these transfers have been performed correctly (yes I know how stupid that sounds).
I would like us not to do that, and I think PMP would be a good option to make sure critical steps are performed correctly.
Here are some Aspiration graphs for 2µl sample
However, because incoming samples might have different DNA concentrations, viscosity, and more we need to do a period of monitoring so we don’t end up with a lot of false positives (positive for detection of error).
Unfortunately, we cannot get the data behind these graphs without signing an NDA, and due to that I might soon not be able to speak about it - great timing on your question btw!
And question for you,
Are you using the predefined neural nets with slight adjustments to the limits? if you’re making your own, are you starting with a detection and then modifying from there?
At this time we are not using the predefined NN, just storing the data for later. We have their NN models and could test and set thresholds on the dataset after the fact.
However, we do not have access to the actual data, only the graph (of the data) shown in their software.
The first job is to try to overlay all the data and “see what pops out”. After that we may want to do som ML ourselves.
Interesting. Are you just running pLLD or MultiSense?
I’m running MultiSense with an anomaly NN and it seems to generate a lot of data that is visible even through the pressure curve viewer.