“GEOAI Machinist”: A Serious Game to Teach the Application of Convolutional Neural Networks to Land Use and Land Cover Classification

Abstract

The rapid growth of the Earth Observation (EO) sector due to advances in technologies such as big data and artificial intelligence increases the demand for skilled workers and, consequently, for tools to support the training of future professionals. Among the services provided by the EO sector, land use and land cover analysis contributes to important decision-making. In this context, “GeoAI Machinist” is a serious educational game in the field of geospatial artificial intelligence that was designed and implemented using a pre-trained land cover and land use classifier. It covers topics relevant to students from higher education in geospatial technologies. As a result, “GeoAI Machinist” serves as a supplemental learning activity, utilizing the educational benefits of serious games as well as the availability of online distribution. The benefits have been assessed in a one-group pretest-posttest design experiment that demonstrated improvements in knowledge and perception of knowledge, as well as positive perceived learning.