Not Only Robots

Whilst we can focus on field robots as an obvious Use Case for AI solutions trained on synthetic datasets, this is not the only market. It is a large market though extending into all parts of agriculture.

Reverse phenotyping

We can take real images and ‘reverse’ them into synthetic plants. This means that researchers can study the development of plants after the plots and field has gone.

Small Plot Trial assessments

Ah, those were the days. Standing in a crop assessing levels of Rhynchosporium in an Irish field (best place for Rhynco) and getting soaked in the soft rain. Back at the office realising that Replicate 3 had a very high variance and then planning a revisit and re-assessment. Did we do that after lunch? There is a better way. Train the AI solution and use a simple vehicle or even a hand-held camera to assess the plots. No bias. Minimised variance. Optimum statistical analysis.


Assessing the growth of a plant. Rather than just imaging and trying to assess biomass, we can train a solution that assesses the variance in growth stage, leaf area and a myriad of other data point in crops. This can aid the development of crop modelling which is a key research area at present.

Crop assessment

Seed companies and plant breeder spend a lot of time looking at crops and assessing their development. We can ensure that every plant and leaf is identified and its position then noted. The AI solution will be able to see and document differences that we never knew existed.

Leave a Comment