Heliostats could represent up to 60% of the investment cost for solar tower plants with more than 100 MWe of power, then reducing the cost of heliostats would have an important cost reduction of the plant. The Heliosun project approaches this cost reduction from three different perspectives. Firstly, an artificial vision system with object recognition is proposed, which allows the closed-loop tracking control of the heliostats. This system, consisting of the installation of a low-cost camera and processor in each one of the heliostats, will eliminate the positioning sensors and improve the tracking accuracy of heliostat, improving the concentrated solar radiation distribution on the solar receiver surface. Moreover, a correct measurement of the atmospheric attenuation suffered by the solar radiation concentrated by the heliostats on its way to the solar receiver, with distances greater than 1500m in large plants, will allow firstly, to perform an adequate selection of those sites with the best characteristics for the deployment of solar tower plants and to optimize the routine operation of the solar plant. Finally, a ray-tracing simulation software, based on OTSun, is intended to be developed, including a more accurate prediction of the behaviour of a solar tower plant with central receiver considering spectral analysis, as well as including all the experimental results presented above. These three approaches will allow to improve the operation of solar tower plants as a whole, optimizing the operation of the solar receiver and the solar field, increasing the technical and economic efficiency of these systems.