Yttria stabilised zirconia (YSZ) is still the most commonly used coating material in thermal barrier coatings (TBC) due to its low thermal conductivity and good mechanical properties; however, it experiences an undesired phase transformation operating at temperatures above 1200 °C and presents poor CMAS resistance, leading to damage and spallation of the coatings. Rare-earth (RE) zirconates, with formula A2Zr2O7, are promising candidates to replace YSZ, as they present great CMAS resistance and superior thermal insulating performance; however, these have low coefficient of thermal expansion and poor mechanical properties. Entropy-stabilization can be a novel strategy to tailor these new TBC materials. Replacing the A-site of the zirconate structure with rare-earth ions (e.g., Yb3+, Nd3+, Sm3+, Eu3+, Gd3+, etc.) in different proportions leads to a highly disordered and distorted structure, improving the required properties to use these zirconates as TBC materials. In this work, several multicomponent rare earth zirconates have been synthesised from a solution precursor route through a new materials discovery apparatus (MDA), which mix the different precursor in the correct amount to obtain the desired ceramic composition. Initially, different RE-zirconates binary systems containing La, Gd and Yb were characterised in terms of CMAS resistance and thermal conductivity. Structural characteristics and physical properties were feed to a machine learning algorithm that suggest new compositions substituting different rare-earth ions. The complex multicomponent rare earth zirconates were produced and tested, and the results were fed to the algorithm all over again. The materials discovery apparatus and the compositions it produces, create a closed loop to further enhance the machine learning model and engineer new rare earth zirconates with enhanced characteristics.