The search and discovery of food sources in central place foragers and bumblebees (genus Bombus) is an instrumental complex problem. Their foraging behaviour can be divided into two distinct phases: exploration and exploitation flights. These two phases are characterised by unique flight behaviours and trajectory patterns. Exploratory flight trajectories are curvier and appear as digressive looping flights in undiscovered areas around the nest without stops at previously visited locations. This serves the purpose of orientating the bumblebee around the nest, learning where the nest is located in relation to its surroundings and exploring possible food sources. In the gradual process of switching from exploration to exploitation, the bumblebee no longer flies multiple loops around the nest, but instead flies directly in single loops to stop at specific locations that have been memorised as food sources. Nonetheless, there is a high degree of inter-individual variation among bumblebee flight patterns. Our aim is to investigate the conditions under which looping behaviour emerges in bumblebees (Bombus terrestris in particular) during exploratory flights in nature. To do this, we use agent-based models in which agents are imbued with an artificial neural network representing the bumblebee brain, combined with an evolutionary process in which thousands of agents evolve over multiple generations. The bumblebee-agents with the best abilities are transferred to the next generation. Our model explores the necessary information needed to reproduce the exploratory behaviour empirically observed, taking into account the location of the nest and the memory of previously explored areas. We expect this approach to be sufficient to recreate the natural exploratory behaviour of bumblebees, characterised by an incremental expansion of the foraging area while maintaining the ability to efficiently return to the nest. This framework analyses the criteria that are being optimised during exploratory behaviour and the cognitive inputs required to achieve this.