Satellite imagery will be an integral part of a bespoke algorithm, developed to predict vining pea yields at harvest.

‘PeaSat’ is a new project, centred on a collaboration between ADAS and the pea grower cooperative, HMC Peas.

The system will utilise satellite imagery to develop a vining pea yield prediction algorithm for farmers.

For vining pea growers and processors, the vital task of harvesting peas at the correct maturity and transporting to factory as quickly as possible thereafter, is logistically challenging at best.

With no way to provide accurate early forecasts of yield, growers and processors often struggle to coordinate throughout the harvesting period, resulting in reduced efficiency and a greater risk of bypass and crop abandonment.

Pea yields

The new ‘PeaSat’ project builds on previous research, which uncovered strong correlations between vining pea flowering and final yield.

With satellite imagery found to be the most efficient way of measuring these correlations, remote sensing specialists at ADAS will examine if satellite data taken at flowering could be used to accurately predict final pea yield. 

It is hoped the resulting algorithm will be ready for use by growers before the 2024 vining pea harvest.

Principal remote sensing consultant with ADAS and project lead, Ben Hockridge, said: “This project has huge potential to streamline processing, and make vining pea harvest easier to plan for and predict.

“By creating an algorithm using satellite imagery, we can provide real-time, up-to-date information for everyone involved, no matter where they are in the process.

“We will be working with technologies that can deliver a number of different image-based resolutions in the field.”

Pea data

HMC Peas, which has had yield mapping viners since 2017, is supporting ADAS researchers by sharing its repository of past yield map data for multiple vining pea varieties.

This data will be used to calibrate the algorithm and ensure accuracy in time for harvest 2024.

Already, grower groups and processors have expressed support for the project.

The theory behind the development of the new technology is as follows – if it is possible to predict the yield of peas coming out of a field, it should be possible to ensure that crops are taken at the required tenderometer reading (TR).

Waste levels will also be kept to a minimum.

The general manager of HMC Peas, Allen Giles said: “If we can accurately predict yields from individual fields, then this could revolutionise the way we harvest vining peas.

“The efficiencies that this can bring will not only increase our productivity, but will also potentially reduce our environmental impact significantly.”