Accessing instrument data is an economics problem
In the past I've written that collecting instrument data is mostly a technical problem, compared to collecting metadata, which is a mix of technical and process/operational. But as I was trying to come up with an eye catching title for this post, I realized that there’s actually more to collecting instrument data. But it isn’t process/operational - it has to do with the economics of selling instruments. So that's what this post will be about. If you're following along, this is part of my ongoing series on the use cases that biotech software needs to support.
There are many things you might want to do with the data that comes from your lab instruments. Some users will prefer to analyze it on the instrument, when possible, because that's the simplest option. Others might want to transfer it straight to their laptop, the next simplest option. But if you're the type of person who reads this newsletter, there's a good chance you'll want to transfer it to a robust, shared storage system, probably in the Cloud. This last one is the most complicated option, but the one that leads to the best long-term outcomes.
There are a few instruments out there that are too simple to justify adding a network adapter. A USB port or equivalent is enough. But many lab instruments these days already have computers inside them, which usually includes a network adapter. Once it's connected to the Internet, it should be easy to automatically upload all your data to exactly where you want it.
Right?
And yet, many manufacturers make this much much harder than it should be. Sure, not *all* manufacturers. There are a few that make it easy, but only a few. The rest do things like making exporting data a ten-step manual process that you have to do multiple times for different parts of the data. They discourage you from adding upload scripts directly to the computer, either with technical obstacles or threatening to void the warranty if you try. They do everything they can to push you towards keeping your data on the instrument or within a software platform that they provide.
I think there are three main reasons that instrument manufacturers (not *all* manufactures) do this. Here they are, ordered from most altruistic to most cynical:
Historically, pretty much all lab users fell into one of those first two buckets of users who want to do the analysis on the instrument or transfer the data to their laptop with a USB stick. So they make things really easy for those users, and don’t give much thought to that third (historically rare) bucket of users.
Instruments are complicated, the software needs to work exactly as expected, and manufacturers don’t trust users to not break everything. Even if your upload agent doesn’t accidentally overwrite files or crash the system, it could easily take up enough memory and CPU time to slow the system down and cause issues. They don’t want to spend their time fielding tech support calls for problems that your IT department caused.
There’s a widespread perception that selling software is a higher margin business than selling hardware. Whether or not this is true, manufacturers (and their VCs) believe they can make more money from selling you software alongside their instruments. And the best way to do that is to sell licenses to the software that is the only way to analyze the data from their instruments.
Now, I try not to use this newsletter to make wild accusations about companies that I will probably want to work with at some point. And truthfully, I have no way to know which of these reasons motivate any specific manufacturer. But I suspect that number three plays some kind of role at least some of the time.
So maybe those problems getting data off your instrument are more of an economics problem than a technical problem.