DESCRIZIONE
Autonomous aquaculture fish feeding system comprising an autonomous vessel (2, 71) having a plurality of temperature and salinity sensors (72) for two or more predetermined pond water depth ranges; an electronic data processor (1) configured for calculating a fish weight prediction model from collected temperature and salinity data from said temperature and salinity sensors (72) from each predetermined pond water depth range; wherein said fish weight prediction model is segmented into one or more prediction model segments, optionally for each of the predetermined pond water depth ranges. The electronic data processor (1) may be configured for calculating the fish weight prediction model, for each prediction model segment, by: accumulating weight gained into an accumulated total weight, wherein the weight gained is calculated from the energy used at each prediction period, calculating the energy used from the accumulated total feeding intake; accumulating feeding intake into accumulated total feeding intake, wherein the feeding intake is calculated from the total weight, a temperature factor and a salinity factor.