A framework for predicting soft-fruit yields and phenology using embedded, networked microsensors, coupled weather models and machine-learning techniques

Abstract

Predicting harvest timing is a key challenge to sustainably develop soft fruit farming and reduce food waste. Soft fruits are perishable, high-value and seasonal, and sales prices are typically time-sensitive.In addition, fruit harvesting is labour-intensive and increasingly expensive making accurate phenological predictions valuable for growers. A novel approach for predicting soft fruit phenology and yields was developed and tested, using strawberries as the model crop. Seedlings were planted in polytunnels, and environmental and yield data were collected throughout the growing season. Over 1.2 million datapoints were collected by networked microsensors which measured spatial and temporal variability in air temperature, relative humidity (RH), soil moisture and photosynthetically active radiation (PAR). Fleeces were added to a subset of the plants to generate additional withinpolytunnel variation. Cumulative fruit yields followed logistic growth curves and the coefficients of these curves were dependent on micro-climatic growing conditions. After 10,000 iterations, machine learning revealed that RH was the optimal factor informing the coefficients of these curves, perhaps because it is an integrative metric of air temperature and water status. Trigonometric models transformed weather forecasts, which showed a relatively low agreement with polytunnel air temperature (R 2 = 0.6) and RH (R 2 = 0.5) measurements, into more accurate polytunnel-specific predictions for temperature and RH (both R 2 = 0.8). We present a framework for using machinelearning techniques to calculate curve coefficients and parametrise coupled weather models which can predict fruit yields and timing to a greater degree of accuracy that previously possible. Dataloggers measuring environmental and yield data could infer model parameters using iterative training for novel fruit varieties or crop types growing in different locations without a-priori phenological information. At this stage in the development of artificial intelligence and networked microsensors, this is a step forward in generating bespoke phenological prediction models to inform and support growers.

Translational challenges and opportunities in biofilm science: a BRIEF for the future

“…Fruit production at this scale requires the engineered optimisation of multiple inputs in climatecontrolled conditions, with parameters such as soil moisture and temperature, air humidity and temperature, pH, macro-and micronutrient levels all closely monitored in order to optimise yield and timed inflorescence (fruiting) while minimising labour. Bespoke subscription-based weather forecasting, soil quality surveys, and remote-sensing data, e.g., hyperspectral imaging, may also be used 31 . Additionally, costly resources are expended on managing these through heating, cooling, irrigating, fertilising, or treating crops 27 .…”

A deep‐learning‐based experiment for benchmarking the performance of global terrestrial vegetation phenology models

“…The methods of machine learning, particularly deep learning, nowadays offer a new means for processing and utilizing remote sensing data. Machine learning‐based studies have emerged on extracting and modelling vegetation LAI with remote sensing data on a regional scale (Almeida et al, 2014; Belda et al., 2020; Czernecki et al., 2018; Dai et al., 2019; Houborg & McCabe, 2018; Lee et al., 2020; Pearson et al., 2020; Xin, Li et al., 2020). Compared to traditional machine learning models, deep learning models have advantages in data mining and modelling because the deep learning models are able to extract multi‐level features from images and/or time series data.…”

Weather Based Strawberry Yield Forecasts at Field Scale Using Statistical and Machine Learning Models

“…Moreover, limited studies have explored the prediction capabilities using random forest. With regard to strawberries, very few machine-learning models have been developed to predict strawberry yield [32,37]. There is a lack of robust and integrated models that couple both field measurements and environmental parameters to make short-term predictions at the field scale using a lower amount of information.Most of the past studies have concentrated on strawberry yield forecasts using meteorological information at the regional scale.…”