When the lab stack meets a dirty bench

by Tom BergLab Systems Editor, HMND

The trouble starts at the handoff. A sample leaves the centrifuge, the instrument spits out a file the vendor swore was standard, somebody copies a value into a shadow spreadsheet because the ELN field does not fit the assay, and now the tidy story in the demo has to survive a Tuesday afternoon with real people in it.

The part vendors keep sanding smooth

LIMS, ELN, and automation software are built for different jobs, and the failure shows up fast when a lab asks one system to behave like the other. A LIMS is the better fit when the work is sample tracking, chain of custody, workflow routing, QC, and instrument integration at scale, while an ELN is built to capture scientific notes, methods, and observations.

That split looks neat in a product deck. On a bench, the chemistry of the day does whatever it wants.

A scientist wants to record an exception. An analyst wants to release results. The system wants every field completed in the right order, with the right metadata, from an instrument export that may still arrive as CSV, PDF, or whatever the vendor left behind in 2014. When the integration only works for the happy path, people build their own path around it.

The workaround becomes the workflow

I have seen this before in migration projects. The old ELN had notes nobody wanted to lose, the new LIMS had the sample IDs, and the instrument software had its own little kingdom with its own naming rules, so the lab carried all three for months and called that transition. By the time the team admitted the truth, the real process lived in three screens, two exports, and a spreadsheet that one senior scientist guarded like the only good wrench in the drawer.

That is where data integrity starts to fray. Manual reentry, duplicate records, and disconnected systems create silos and errors, and ELN based workflows in particular tend to push scientists back toward notebooks, personal files, and spreadsheets when the software does not fit the work.

The vendor will call that flexibility. The lab calls it overtime with a grin that does not reach the eyes.

Scaling makes the ugly parts louder

Current pressure on lab data infrastructure is not coming from one hero system. It is coming from more instruments, more file types, more regulated handoffs, and more pressure to connect ELN, LIMS, SDMS, and automation without dropping the chain of custody or slowing the bench. Integration is where the floorboards creak, because every extra export or manual transfer creates another place for sample tracking, assay metadata, and result provenance to drift apart.

A clean API helps. So does a sane data model. But the real test is whether a system can absorb a messy human workaround without turning it into permanent process debt.

That is the part the glossy architecture slides skip. Daniel would call half of these incidents an API problem, and he would be right, but the deeper issue is usually workflow design, not just transport. If the assay metadata lives in one place, the sample status in another, and the instrument export in a third, the lab is already doing reconciliation work whether anyone admitted it or not.

What breaks first

The first thing to break is usually trust. A bench scientist stops believing the ELN is the source of record because it cannot hold the awkward exception, or the LIMS loses credibility because it cannot keep up with the instrument reality and the sample queue becomes a pile of partial states.

After that comes the familiar mess. Someone keeps a spreadsheet. Someone else pastes a file link into a chat thread. The migration team says the cutover is done while the lab keeps one foot in the old system because the old system still knows how to survive the oddball sample and the instrument that exports a cursed filename at 4:17 p.m.

If you want a practical question for your next vendor meeting, ask what happens when the sample arrives mislabeled, the assay metadata changes after run start, and the instrument export does not match the template. The answer tells you more than the roadmap does.

If this handoff problem is sitting on your desk, write hello@example.com. We build the systems side of that mess.