%We envision this concept can apply to many other objects in our environment.
%The question we must always keep in mind is the degree of control we would like to keep over these wandering TUIs.
-\newpage
\section{Conclusion}
-Regard critique sur les projets :
-
-- activibe : vocabulaire limité et design des sets pas complètement propre. Par contre l'étude longitudinale est rare dans ce domaine et permet une validation écologique
-
-- textures : résultats difficiles à généraliser mais ils ont permis d'améliorer le device. Variabilité du device avec un vrai doigt dessus, dans des conditions réelles.
-
-- printgets
-
-- actuated devices
-
-Ramener au modèle du début et questions de recherche
-
-L'orientation de la recherche dépend beaucoup des personnes avec qui on travaille. textures et printgets partent d'une problématique similaire : améliorer un proto de recherche haptique en vue de commercialisation. Approches différentes : sciences cognitives / audio.
-
-Future work:
-\begin{itemize}
-\item why 1000Hz
-\item Force feedback processor
-\end{itemize}
+The work that was presented in this chapter illustrates the systematic approach in my research.
+The objective is not to focus on a particular technology, a parrticular problem or a particular context.
+Rather, I search for an appropriate technology for a given problem in a given context.
+The type of research questions I address is large, and depends in particular on the expertise of my collaborators and the objectives of the research project.
+For example in the \emph{Activibe} project, I collaborated with other researchers in HCI who were interested in using haptic cues for promoting behavior change.
+The focus was not on designing a new device or even a new haptic technology, but to design a way to encode information with an off-the-shelf device.
+Hence, we worked on the design, implmentation and evaluation of tactons.
+The approach was different with \emph{tactile textures} and \emph{printgets}.
+In these case I collaborated with researchers with a background in engineering who were designing a new haptic technology.
+The two situations were different though.
+In the first case I also worked with a postdoc who had a background in cognitive sciences.
+Therefore our studies focused on the perception and interpretation of tactile textures.
+In the second case I worked with a postdoc who had a background in audio and music technologies.
+As a consequence, our work was directed to the signal generation and authoring, and the technical apparatus.
+Finally, with the \emph{actuated devices} project I mainly worked with HCI colleagues, and a master intern with an electrical engineering background.
+The intern designed and implemented most of the robotics part of Living Desktop, I worked mostly on the hardware part of Métamorphe, the software part of both and the observation study, and all of us worked on user studies and the application scenarios.
+The complementarity of expertise in my research project is an important aspect that not only guides my own contribution, but also the whole direction of the project.
+
+These research project have a few limitations.
+Acknowledging them is not only umportant regarding to these projects, but this is also useful hindsights for the evolution of my research practices on the same type of projects.
+While the longitudinal study we performed with \emph{Activibe} is rare and insightful on this kind of topic, there is room for improvement on the tacton design methodology.
+The design rationale was not systematic, so even if we found a satisfactory solution there is probably many other possible designs.
+The difficulty is that the design space is huge, and there are many point of views.
+The main consequence on the results is the difficulty to generalize these results to bigger sets.
+The design of large tactons sets is hard in general.
+The works of \emph{tactile textures} and \emph{printgets} were interesting examples of projects in which the research on the technology and the research on the usage of the technology feed each other.
+They are illustrations of Huot's \emph{designeering} concept~\cite{huot13}.
+One of the limitations is that research prototypes are built upon simulations.
+However, real world conditions can have a huge impact on the device performance, and the user's perception.
+This is why an iterative process is efficient at taking ito account both the technological and interaction concerns.
+However the other limitation is the slow iteration process due do the long manufactoring of prototypes.
+HCI uses low-fidelity prototypes to bootstrap the first cycles of iterative design.
+It is important in such projects to identify alternative technologies that can be used for prototyping.
+This way, as HCI researchers, we can give our engineering colleagues specifications at the early stage of the project rather than waiting for their early prototypes that already commit to choices that cannot be changed.
+In the \emph{actuated devices} project, we built working prototypes that enabled us to explore the scenarios in real contexts.
+The Métamorphe was robust enough to evaluate interactive properties, such as the users' ability to locate raised keys.
+However technical limitations prevented us from actuating all keys, or add all the sensors we would need to explore some of the new input we imagined, typically rotating the keys.
+The Living Desktop was functional, but with technical limitations on speed or forces.
+Therefore a user evaluation with the prototype could have been biased due to these technical limitation.
+This is why we evaluated the scenarios with videos.
+We would certainly have collected more valuable data with more robust prototypes.
+
+In this part of my research activities, I leveraged the sense of touch for the design of interactive systems.
+Most of this work focused on tactile sensations, but I remained interested in kinesthetic sensations as well.
+The design of force feedback devices is complex due to both the mechanical constraints and signal generation of the device side, and the perception and interpretation of the mechanical effect on the user side (see \reffig{fig:hapticpath}).
+My next projects in this area will cover an evaluation of the effects of the haptic loop frequency on the perception of haptic feedback.
+It will require the design and implementation of a simple but fast force feedback device.
+
+This chapter focused on haptics as the sense of touch, which is the usual point of view on haptics.
+However, as we have seen with the actuated devices in particular, haptic is also about manipulation. This is the subject of the next chapter.
+
+%One of the questions I would like to study is the force feedback loop, and especially its frequency.
+%The haptic loop requires a high frequency for a smooth and realistic experience.
+
+%\cite{goff67}
+%Future work:
+%\begin{itemize}
+%\item why 1000Hz \cite{massie94} \cite{ramstein94,ramstein95} Hapticmaster\cite{vanderlinde02} (2500 Hz)\cite{leonard15}
+%\item Force feedback processor
+%\item what about input? combination?
+%\end{itemize}
+
+%vanderlinde02: Because this frequency is approximately ten times higher than the maximal human discrepancy value [Burdea, 1996], it is assumed to be high enough to guarantee a haptic quality for a smooth and realistic experience
projects / hdr.git / commitdiff