The APPF’s FieldExplorer completed imaging in 16 field trials in its first year of operation and included clients from universities, government research organisations and breeding companies. This ground-based phenotyping platform is able to measure plant traits such as biomass and stress tolerance in field trials non-destructively using a suite of on-board sensors.
The FieldExplorer is based at APPF’s The Plant Accelerator in Adelaide and we talked to Technology and Development Lead, Dr Darren Plett, about how the season went.
What trials did the FieldExplorer complete?
“Trials this year spanned a wide range of research areas, including a ground-truthing project for Elders and D-CAT related to their new satellite-based stock grazing management tool which assesses forage biomass. The management tool uses satellite information to get an indication of biomass availability for grazing in perennial pastures and ground-truthing will help to check that the satellites are picking up accurate information.
“The FieldExplorer was also used in an oaten hay trial funded by Agrifutures to develop methods to use LiDAR and hyperspectral imagery to measure biomass and feed quality traits.
“In other trials, chickpeas were imaged to assess for Ascochyta blight in the field, and to determine if images could be used to automate pod and seed counts as an indication of chickpea susceptibility to chilling, or cold weather at flowering. The aim is to be able to quickly assess susceptibility and identify those lines able to tolerate chilling at a critical time.
“In paddock-scale trials with the CSIRO Future Farms project, barley biomass estimates were obtained with LiDAR data with the goal to determine whether the technology will be suitable to provide farmers with the ability to make decisions on nitrogen management ‘on the fly.”
What does the FieldExplorer bring to these projects?
“Measurements are obtained using imagery, so are non-destructive. This means the client benefits because they can obtain their data faster, more easily and potentially more accurately as compared to traditional ‘wet-lab’ methods which can be laborious and costly.
“As an example, in a trial looking at barley head loss after maturity, we were able to image the trial before maturity and again three or four weeks later and use image analysis techniques to count heads in the plots This is much quicker than manually counting heads and enables assessment of larger numbers of plots, a requirement of programs breeding for improvement of this trait.”
What happens with all the data you have collected?
“Analysis of the terabytes of hyperspectral and LiDAR data is well underway thanks to our team, including our new Image and Data Lead, Dr Ken Clarke, and Postdoctoral Research Fellow, Andrew Hennessy. The team is busy building the required pipelines to analyse the data and translate the information into tailored and useful management decision tools.
What is next Darren?
“Seeing the value of the FieldExplorer in assessing new crop varieties, disease and pest management, nutrition and plant growth, public and commercial organisations are keen to use the machine next season develop the multi-year data sets required to reach their research goals.
“We are also investigating new research areas and are involved in trials in the horticultural industry. Examples include one where we are measuring the effects of a major fungal disease in potatoes and the FieldExplorer will also potentially be imaging grapevines in a nursery and an onion trial for disease symptoms.”
“Another busy year is ahead with trial work as well as upgrades and new solutions to improve the collection and analysis of FieldExplorer data. The machine has plenty of promise and so it is an exciting research space in which to be involved.”