There is provided a method of assessing a characteristic of a plant ( 2 ) , comprising : obtaining a training dataset , wherein the training dataset comprises first data characterising a first electrical signal obtained from a first plant during a first time period when a stressor is present in the first plant or a growth environment of the first plant , second data characterising a second electrical signal obtained from the first plant during a second time period when a stressor is not present in the first plant or the growth environment of the first plant , and third data indicative of a characteristic of the first plant during the first time period and a characteristic of the first plant during the second time period ; training a machine learning model based upon the training dataset ; obtaining a third electrical signal from a second plant ; and assessing , using the trained machine learning model , a characteristic of the second plant based upon the third electrical signal