<p>Three-dimensional (3D) interfaces are made with the ambition to reinvent traditional two-dimensional (2D) displays into intuitive 3D environments that allow natural interactions and guarantee users’ satisfaction. However, reality is far from the expectations and 3D interfaces remain experimental. The results of empirical studies comparing 2D and 3D interfaces are incoherent and do not indisputably endorse the development of 3D applications. The reason behind this incoherency is that empirical comparisons often discard several factors that go beyond the specific style of the interface itself although they play a major role in human performance. Specifically: perceptual factors, factors related to semantics, contextual factors.</p><p>The identification of these factors was carried out through a set of empirical studies tackling two applications domains: information management (namely, the retrieval of data and on spatial memory tasks) and Air Traffic Control. Concerning the first domain, the results suggest that 3D interfaces can support spatial memory if the 3D interface is provided with a spatial structure that has also a semantic function. Also, the specific content of the objects disseminated in the 3D structure seem to affect performance. It is argued that when there is a strategic coupling of the semantics of the spatial structure with the meanings of the objects, 3D interfaces could enhance spatial memory. Concerning the second domain, the results indicate that 3D interfaces could support controllers’ tasks only for a limited set of activities. These do not include standard monitoring tasks where the presence of 3D would actually cause detriment to the performance due to perceptual biases. Finally, it is claimed that the idea of familiarity with an interface style can affect the way people interact with it and, despite the fuzziness of the concept, familiarity may represent a real challenge in users’ acceptance of 3D interfaces. </p>

Three-dimensional (3D) interfaces are made with the ambition to reinvent traditional two-dimensional (2D) displays into intuitive 3D environments that allow natural interactions and guarantee users’ satisfaction. However, reality is far from the expectations and 3D interfaces remain experimental. The results of empirical studies comparing 2D and 3D interfaces are incoherent and do not indisputably endorse the development of 3D applications. The reason behind this incoherency is that empirical comparisons often discard several factors that go beyond the specific style of the interface itself although they play a major role in human performance. Specifically: perceptual factors, factors related to semantics, contextual factors. The identification of these factors was carried out through a set of empirical studies tackling two applications domains: information management (namely, the retrieval of data and on spatial memory tasks) and Air Traffic Control. Concerning the first domain, the results suggest that 3D interfaces can support spatial memory if the 3D interface is provided with a spatial structure that has also a semantic function. Also, the specific content of the objects disseminated in the 3D structure seem to affect performance. It is argued that when there is a strategic coupling of the semantics of the spatial structure with the meanings of the objects, 3D interfaces could enhance spatial memory. Concerning the second domain, the results indicate that 3D interfaces could support controllers’ tasks only for a limited set of activities. These do not include standard monitoring tasks where the presence of 3D would actually cause detriment to the performance due to perceptual biases. Finally, it is claimed that the idea of familiarity with an interface style can affect the way people interact with it and, despite the fuzziness of the concept, familiarity may represent a real challenge in users’ acceptance of 3D interfaces.

Context. 3D visualisations are highly demanded in different industries such as virtual factories. However, the benefits that 3D representations can bring to this industry have not been fully explored, being most of the representations either photorealistic or presenting abstract visualisations. Objectives. This thesis explores and compares two prototypes that present a visualisation of the process state of a factory. The first prototype presents a generic interface in which primitive 3D shapes convey the information of the factory status. The second prototype is complemented with specific and iconic 3D models of the factory that help the users associating the conveyed information to the factory flow. The motivation behind this dissertation is that the type of generic interface presented can lead to more reusable interfaces in the future. Methods. For the creation and development of the prototypes, the user-centered design process was followed in which the designs are iterated with users of the factory. Based on the two prototypes, a usability evaluation is conducted to analyse the perceived usability and the usability performance. This is complemented with post-interviews with all the participants. The results are presented attending to the triangulation methodology to support the strength of the qualitative findings. Conclusions. The results show that both interfaces are perceived as highly usable. However, the 3D iconic interface seemed to help the users more in depicting a better mental model of the factory flow, helping the users to complete most of the tasks with faster times. / This thesis explores and compares two prototypes that present a visualisation of the process state of a factory. The first prototype presents a generic interface in which primitive 3D shapes convey the information of the factory status. The second prototype is complemented with specific and iconic 3D models of the factory that help the users associating the conveyed information to the factory flow. The motivation behind this dissertation is that the type of generic interface presented can lead to more reusable interfaces in the future.