While energy conservation is an urgent need, many people still misuse thermostats to control heating systems, resulting in wasted energy. One of the main causes is the lack of immediate thermal feedback. In this work, we investigate how immediate touch-based thermal feedback can be an appropriate interaction modality to improve the thermostat UI for manual control, and allow anticipation of the outcome of setpoint adjustment through user interaction. Based on two mixed-methods experimental user studies, we demonstrate the applicability of immediate thermal feedback to translate and anticipate a sense of thermal comfort that is meaningful and satisfying to the user. We also show the usability of this feedback as an interaction modality for adjusting a temperature setpoint, which is perceived as simple, natural and accurate.

Modular Tangible User Interfaces (TUIs) –i.e., UIs made of small-scale physical modules– offer novel opportunities for tangible interaction thanks to their highly customizable form factor. Such modular TUIs were proposed with different shape of modules and bonding strength between them. The problem we address in this paper is the lack of knowledge of how bonding strength and shape of the modules impact usability. We present the first study exploring the impact of bonding strength and module shape on subjective user ratings when interacting with a magnetic modular prototype. We assessed three levels of bonding strength (low, mid, high) and two shapes (cubes and rounded cubes) in a controlled user study. Participants performed eight common manipulations found in the literature for (non-)modular TUIs. Experimental results showed that (1) cubic modules are overall easier and more satisfying to manipulate, except for precision and bending tasks, (2) low strength impairs UI solidity, but high strength impairs precision tasks with cubic modules.

Shape-changing User Interfaces (UIs) explore the ability of a UI to change its physical shape to support multiple interaction modalities for users’ input and/or system’s output. An approach currently studied to implement such interfaces at a high resolution is based on mm-sized, round, and self-actuated modules. The problem we tackle in this paper is to find the range of usable sizes of such modules, to better inform the trade-off between usability and technological feasibility. We assessed four sliders in a controlled user study: a standard slider and three sliders made of mock-up rounded modules of 1 mm, 2.5 mm, and 5 mm. Experimental results show that (1) 5 mm modules significantly impair performance for the pursuit task and subjective perception for both tasks, (2) performance increases when the size of modules decreases, but (3) users reportedly enjoyed the haptic feedback provided by 1 mm to 2.5 mm modules. These results provide deeper understanding on the impact of the size of modules on performance and subjective perception to inform current technological development of physical user interfaces made of small robotic modules.

Shape-changing User Interfaces attract growing interest in Human-Computer Interaction. Modular robotics offer a great opportunity for their implementation. However, the current theoretical and technical advances of modular robotics are fragmented and little centered on the user. To unify existing work and center future research on the user, we perform a systematic literature review enabling us to build a unifying space for the design of modular shape-changing user interfaces.Our aim is to bridge the gap between HCI and robotics. We relate properties of different domains and identify inconsistencies to structure the design space. Towards this aim, we conduct a thorough cross-disciplinary survey to propose: 1) a set of design properties at the scale of the interface (macro-scale) and at the scale of the modules (micro-scale) and 2) the impact of these properties on each other. This paper can be used to describe and compare existing modular shape-changing UIs and generate new design ideas by building upon knowledge from robotics and HCI.

This paper contributes a first comparative study of three techniques for selecting 3D objects anchored to the table in tabletop Augmented Reality (AR). The impetus for this study is that touch interaction makes more sense when the targeted objects are anchored to the table. We experimentally compare touch and a mixed (touch+mid-air) techniques with the common direct mid-air technique. The touch and mixed techniques involve a decomposition of the 3D task into a 2D task by touch on the table followed by a 1D task by touch or mid-air interaction. Results show that: (1) The touch and mixed techniques present completion times similar to the mid-air technique and are more accurate than the mid-air technique; (2) The mixed technique defines a good compromise between accuracy of touch interaction and speed of mid-air interaction.

We experimentally compare the performance and usability of tablet-based and see-through head-mounted display (HMD)-based interaction techniques for selecting 3D virtual objects projected on a table. This study is a first step toward a better understanding of the advantages and limitations of these interaction techniques, with the perspective of improving interaction with augmented maps. To this end, we evaluate the performance of 3 interaction techniques in selecting 3D virtual objects in sparse and dense environments: (1) the direct touch interaction with a HMD; (2) the ray-casting interaction with a HMD; and (3) the touch interaction on a tablet. Our results show that the two techniques using a HMD are faster, less physically tiring and preferred by the participants over the tablet. The HMD-based interaction techniques perform equally well but the direct touch technique seems to be less impacted by small targets and occlusion.

The exploration and discovery of nature through walking, cycling, touring, trekking and hiking are targeted by this work, part of an on-going research on mobile-based persuasive user interaction. Considering HCI solutions dedicated to discover and explore nature, we uncover the concept of persuasive event and several classes of roles that users may endorse during a long-term usage of such interactive solutions. We claim the importance of these events and roles as, leveraging persuasion, they may contribute to improve nature discovery experience and to sustain motivation through user interaction. In this work we share these insights from our studies on the Mhikes platform, designed for such kind of nature-involved activities.

In this paper we present the Clover architectural model, a new conceptual architectural model for groupware. Our model results from the combination of the layer approach of Dewan's generic architecture with the functional decomposition of the Clover design model. The Clover design model defines three classes of services that a groupware application may support, namely, production, communication and coordination services. The three classes of services can be found in each functional layer of our model. Our model is illustrated with a working system, the CoVitesse system, its software being organized according to our Clover architectural model.