What should labs know about SphereFlash AI Colony Counting Software from IUL?

Updated

SphereFlash is IUL's AI-powered colony counting software, positioned as a way to automate the plate-reading step in microbiological quality control. It is worth evaluating if your lab counts plates routinely and wants to test whether automation fits your organisms, plate types, and throughput. IUL lists an "Activate Free Trial" entry point on its site, so you can assess it on your own plates before committing. It is not a substitute for your own validation and compliance review.

Where SphereFlash sits in IUL's product line

IUL has operated since 1987 and now runs two main divisions: microbiology lab solutions (IUL Micro) and lateral flow readers (MedIUL). SphereFlash belongs to the microbiology side, alongside hardware such as the Spin Air air sampler and the iPeak lateral flow reader family.

That matters for a buying decision in two ways:

  • It is software, not a standalone instrument. You are adding an analysis layer to plates you already prepare and incubate, not replacing your incubation or sampling workflow.
  • It is one part of a broader lab portfolio. If you are already considering IUL hardware, the software may integrate into a single-vendor setup; if not, judge it on its own merits.

What the AI colony counting actually changes

Manual colony counting is repetitive, operator-dependent, and hard to audit at scale. Automated counting targets that step specifically:

  • Consistency — the same plate image is evaluated by the same rules each time, reducing reader-to-reader variation.
  • Throughput — plates can be processed without an analyst sitting at the bench for each one.
  • Traceability — digital counts and images are easier to store and review than handwritten tallies.

The limits matter just as much. IUL's own event description for its Lunch & Learn workshop asks "what are the limitations of today's systems?" — a fair signal that AI counting is not universally accurate. Expect edge cases with:

  • Crowded or confluent plates where colonies merge
  • Chromogenic or colored media that shift contrast
  • Very small colonies, swarming organisms, or heavy debris
  • Unusual plate formats or lighting conditions

How to evaluate it before buying

Use the free trial as a structured test rather than a quick look. A practical sequence:

  1. Activate the trial through the link on IUL's site.
  2. Collect a representative plate set — include your routine organisms, your typical dilutions, and deliberately difficult plates (crowded, chromogenic, small colonies).
  3. Count the same plates manually and record the reference values.
  4. Compare automated versus manual counts and note where they diverge.
  5. Repeat across operators and days to see whether results are stable.
  6. Document the comparison — this becomes the basis for your validation file.

The trial tells you whether the software handles your plates. Vendor demos on clean, well-separated colonies rarely reflect a real lab's worst samples.

Questions to settle before you commit

Question Why it matters
Which plate types and sizes do you use? Confirm the software supports your formats, not just standard Petri dishes
Which organisms and media dominate? Chromogenic and selective media are common failure points for image-based counting
What is your daily plate volume? Automation pays off at volume; low counts may not justify the change
What does your accreditation require? You will likely need documented validation and a defined procedure for discrepant counts
How do results export? Check integration with your LIMS or record-keeping before, not after, purchase
What is the pricing and licensing model? The site shows a free trial but no published pricing — ask directly

Seeing it in context

IUL runs events where the software is demonstrated live, including a Lunch & Learn workshop on AI-powered colony counting (14 October 2026, Leonberg, Germany) and a presence at ARABLAB 2026 in Dubai. If you are close to a decision, a live demo lets you ask the limitation questions directly rather than inferring them from marketing material.

Bottom line

SphereFlash is a reasonable candidate if your lab does regular colony counting and you want to test AI-assisted reading on your own plates at no initial cost through the trial. Treat the trial as a validation exercise, not a demo: build a plate set that includes your hardest samples, compare against manual counts, and confirm plate-type, organism, and export requirements with IUL before assuming fit. If your plates are mostly crowded, chromogenic, or low-volume, test carefully — that is where automated counting most often falls short.