In microbiology laboratories, plate counting has long been a necessary but time-intensive part of daily work. While essential to food safety and quality control, manual reading and recording can consume valuable technician time that could otherwise be spent on higher-level analysis and decision-making.
That challenge is what helped inspire the development of Compact Vision — AFD’s flagship intelligent plate reading system designed to automate colony counting while strengthening traceability, workflow consistency, and data visibility. For laboratories using Compact Dry, Compact Vision brings a new level of ease and consistency to daily plate reading.
To explore how the technology has evolved and where laboratory automation is heading next, we spoke with the developer behind Compact Vision about machine learning, lab workflows, and why the future of microbiology still depends on skilled technicians.
Building More Than a Plate Counter
According to the development team, Compact Vision’s success came from integrating multiple technologies intentionally around the real needs of laboratory users.
“Compact Vision is not about one specific scientific breakthrough. It’s about putting different pieces together and making a system work for the specific needs of the lab technician.”
The system combines cloud infrastructure, embedded computing, machine learning, and controlled imaging conditions to create a practical tool that fits naturally into laboratory routines.
Early prototypes focused on a core technical question: could a camera-based system reliably identify microorganisms on plates inside a controlled environment? But as customer conversations expanded, the scope changed.
What emerged was something broader than automated counting.
From Image Recognition to Workflow Infrastructure
One recurring insight from early users was that plate counting was only one part of a much larger workflow.
Technicians were also manually documenting plates, maintaining records, and often capturing images independently to preserve traceability.
“It became very clear that this was not only a technical problem — it was a system-of-record problem.”
That realization pushed development toward a larger objective: not only identifying colonies accurately, but creating a digital workflow that improves how labs manage information.
Today, Compact Vision supports traceability, image capture, data organization, and long-term recordkeeping in addition to colony interpretation.
Why Microbiological Plate Reading Is Harder Than It Looks
Automated plate interpretation sounds straightforward until real-world variability enters the picture.
Different Compact Dry plate types behave differently. Incubation conditions shift. Even highly controlled laboratory environments still produce significant visual variation.
Traditional rule-based computer vision systems struggle in that context.
“Much of what lab technicians do is implicit and intuitive. You would be hard pressed to write down every rule they follow.”
That is where machine learning becomes valuable.
Rather than forcing rigid rule sets, Compact Vision is trained to recognize patterns in ways that more closely mirror human interpretation while maintaining consistency under controlled imaging conditions.
Why Human Expertise Still Matters
One theme that came up repeatedly during the conversation was that automation should not be viewed as replacement.
It should be viewed as leverage.
“Compact Vision is not a replacement for a lab technician — it is an enhancer.”
Rather than spending time on repetitive counting tasks, technicians can focus more attention on coordination, interpretation, troubleshooting, and decision-making.
In fact, as labs adopt digital tools, skilled human judgment becomes even more important.
The goal is not to remove technicians from the process, but to strengthen what they can accomplish.
“I hope users feel like it gives them superpowers.”
Customer Feedback as a Product Driver
One of the strongest themes behind Compact Vision’s evolution has been customer collaboration.
Labs frequently surface edge cases — unusual products, uncommon growth patterns, or workflow needs that were not obvious during development.
Those cases directly inform the roadmap.
“Customer feedback is central to the product roadmap. This is not about making technology — it is about solving customer problems.”
That feedback loop allows data to be incorporated into machine learning updates while also shaping interface improvements and feature priorities.
In several early deployments, simplicity emerged as a critical lesson.
More functionality did not always improve usability.
Small interface changes often had major impact on adoption.
Looking Beyond Plate Reading
While Compact Vision is currently known for automated colony counting, the longer-term view is broader.
The system increasingly acts as a digital sensor inside the lab — one capable of generating structured information that can support traceability, records, and future workflow integration.
“The possibilities go well beyond plate reading.”
For laboratories, that opens new opportunities:
· stronger consistency across technicians
· faster data retrieval
· easier audit readiness
· better long-term process visibility
What Success Looks Like for Laboratories
For many laboratories already using Compact Dry for routine enumeration and pathogen screening, digitizing interpretation creates a natural next step toward a more connected workflow.
When asked what success looks like after implementation, the answer was simple:
“When the bottleneck stops being how many plates can be analyzed in a certain amount of time.”
That is where ROI begins — not simply in faster counting, but in freeing skilled personnel to contribute where they add the most value.
For AFD, that aligns closely with how we view technology adoption: not as equipment alone, but as part of a broader laboratory partnership.
Because the best automation does not just count better.
It helps laboratories work better.
Interested in seeing Compact Vision in action?
AFD works closely with laboratories to understand workflow needs, implementation goals, and validation requirements before deployment.
If you would like to learn more about how Compact Vision fits into your laboratory workflow, our team would be glad to connect.
