Drones are becoming an ever increasing part of our lives. From airshows to surveillance systems, their use varies greatly and, as technology improves, this variety will continue to increase. As drones get smaller, new applications arise. For example, drones for monitoring crops or searching for survivors in hard to reach places in disaster areas. One of the main fields of improvement to drone design is energy efficiency. The more energy efficient a drone is, the longer it can fly and the more it can carry. This is the objective of the Atalanta project. It seeks to develop a drone with a wingspan smaller than 15 cm and an improved energy efficiency over other drones. It does this by two means. Firstly, it uses a bio-inspired flapping wing solution, mimicking the methods used by insects to fly. This carries a greater efficiency over the more ubiquitous rotary wing design. Secondly, it looks to apply resonance, allowing more energy conservation between wing flaps. In its current state, the Atalanta project has gathered an expansive amount of theoretical knowledge, but lacks the validation a prototype would provide. This report gathered the knowledge contained within the Atalanta project and supplemented it with new information generated in the last two years. It identified three main challenges to overcome and provided a set of guidelines to overcome them. The challenges are as follows: Firstly, the total design is too heavy for the theoretical lift generated by the wings. Secondly, the wings, in their current form, are too heavy for an actuator to operate at the right frequency. Thirdly, little documentation exists on the design of the resonant element the design has to be built around. If all three of these challenges are addressed, it should be possible to create a flying prototype for the Atalanta project using an external power source and external control. To overcome the limits of the generated lift, varying actuators were evaluated. As no actuator was found meeting the projects requirements, a gearbox solution was sought instead. Through the use of a linear frequency divider, the available power from an actuator could be doubled by doubling the operating frequency of the actuator without changing the flapping frequency of a wing. A linear frequency divider was developed for this purpose. As this element also fulfils the role of the Atalanta body, the third challenge was also addressed through this element by designing it to function as a resonant element, vibrating at the desired frequency. This linear frequency divider was successfully developed, but was unable to be integrated into the prototype within the available time. To address the weight of the wings, a vacuum forming method was investigated. By applying heat to a film of plastic and pulling it over a mold using a vacuum, a wing of a desired shape can be created. By introducing localised folds in this wing, it can be reinforced and be made stiffer. Although this seemed promising, no suitably light wing was produced using this method, but further research is likely to produce positive results. As only mylar was investigated in this paper, evaluation of other materials could prove productive. Alternatively, using a combination of vacuum forming and local reinforcements may well create a near rigid wing. Despite the increase in losses due to the heavier than desired wings, the drone presented in this paper could theoretically fly for short periods of time, as the losses of the flight are compensated for by the energy provided by the used actuator. It was originally predicted that the Atalanta would have a power density requirement of 30 W/kg. Despite the heavier wings and the increase in weight from the components added by the linear frequency divider, the new loss to mass ratio is only 38.4 W/kg. The weight of the prototype has increased as well, however. Despite all this, the actuator is predicted to be able to deliver 1.15 watts of power with the losses being 0.713 watts. As such, with the addition of the linear frequency divider, the prototype can theoretically fly, if integrated properly. Also of note is the lack of optimisation for the linear frequency divider. As this is the first paper investigating this part, a substantial amount of time was spent evaluating the effectiveness of the part. As such, further investigation in this area could lead to further optimisation or even the inclusion of a second divider, allowing the frequency of the actuator, therefore its power, to be doubled again.