Australian Plant Phenomics Facility#
The Australian Plant Phenomics Facility is a world-leading research infrastructure focused on developing new and improved crops, healthier food, and more sustainable agricultural practices. Their high-throughput experiments capture thousands of plant images daily, which are automatically processed to extract growth measurements. However, the team had no efficient way to see the bigger picture across these large image datasets — making it difficult to validate results, spot problems, or identify trends.
We deployed Zegami to give researchers the ability to visually explore entire experiments at a glance. By viewing thousands of images simultaneously and filtering by variables such as treatment, species, and time, researchers could quickly identify outliers and patterns that would have been impossible to catch by looking at images individually or relying on automated analysis alone.
This proved its value almost immediately. In one project, growth charts showed an unexpected dip that appeared to be a biological response. Visual inspection through Zegami revealed the real cause: a power outage had changed the time of day images were taken, altering lighting conditions and throwing off the image segmentation. In another case, what looked like a plant nearly doubling in size turned out to be an RFID reading error where two plants had been imaged simultaneously.
These kinds of subtle problems can render days of experimental results unusable. By catching them early, Zegami delivered significant cost savings and became an integral part of APPF’s workflow for data exploration, quality control, and collaboration with research partners.