compositional controller synthesis for hybrid systems
Axis : DataSense
Subject : Interface sensorimotrice pour environnements virtuels collaboratifs basés sur des dispositifs interactifs hétérogènes : application au design industriel
Directors : Patrick Bourdot, LIMSI et Cédric Fleury, LRI
Institution :LIMSI, LRI
Administrator Laboratory : LIMSI
PhD Student : Yujiro OKUYA
Beginning : 2015
Scientific production :
- Y. Okuya, N. Ladeveze, O. Gladin, C. Fleury and P. Bourdot. ”Distributed Architecture for Remote Collaborative Modification of Parametric CAD Data”, In 3DCVE Workshop, IEEEVR2018, Germany 2018.
- P. Martin, S. Masfrand, Y. Okuya and P. Bourdot. "A VR-CAD Data Model for Immersive Design." International Conference on Augmented Reality, Virtual Reality and Computer Graphics. Springer, 2017.
This thesis project has as its scientific problem the development of a sensorimotor interface model capable of versatility on heterogeneous interactive devices, mainly based on large screen interactive visualization systems (such as the Wall of Images) and immersive rooms (such as CAVE1).
The context of this study is the growing need of industrialists, for the design and review of project modifications, for effective collaborative interactions in large scale visual and/or immersive devices, which alone allow scale 1 activities on complex digital models of their projects. These new working modes (large device, immersion, collaboration) allow productivity gains in product design, but the interaction with 3D content in these types of devices still lacks much too much precision, and a fortiori when it targets collaborative phases (co-manipulation, and especially, co-modification of shapes).
The purpose of this thesis is therefore to study the Haptic and Pseudo-Haptic approaches to solve this problem, and more generally to develop a sensorimotor model that can be adapted according to the technologies available on the different interactive interconnected devices in the collaborative context of interest to us, namely collaborative design and the multi-site modification review of projects.
This thesis topic naturally fits in with the research themes of the DIGISCOPE team, whose two bearers and their respective teams are partners.
Scientific Objective :
Apart from assembly simulation, the equivalence between Haptics and Pseudo-Haptics is still a vast research project.
The first scientific objective of this thesis is to determine the means to bring the two approaches to be perceived as almost equivalent for the aid to the modification of complex 3D shapes, in the spirit of establishing a certain interchangeability between them. The equivalence sought will not necessarily be in the register of "realism", which may be either difficult to achieve at the technological level or not adapted to human perception. As a result, we will not preclude ourselves from considering different interactive paradigms depending on whether the rendering is haptic or pseudo-haptic, the goal being to reach comparable levels of precision, performance and cognition for the complex tasks in question. Compared to gestural interactions without assistance, the difficulty is not only to bring a gain in precision that is perceptually and cognitively comparable in these two modes of rendering, but also to guarantee this equivalence in the combination of sensorimotor assistance (coincidence and/or sequence of elementary feelings) that complex interactions may require.
The second scientific objective is obviously to demonstrate the contributions of this approach for collaborative design in heterogeneous interactive devices which from the point of view of industrial needs are remote. Here again the question of haptic / pseudo-haptic equivalence will be studied, but this time from the point of view of the interoperability of these approaches. One of the key issues will be to integrate the delay component into this sensorimotor interface model for collaborative virtual environments, due to the latencies that remote communication between interactive devices can introduce. Upstream, of course, there will be questions of distributed architecture and data synchronization between devices.
The expected benefits mainly concern the improved usability of collaborative virtual environments based on heterogeneous interactive devices, equipment typically targeted by the Equipex DIGISCOPE research program. Developing a generic sensorimotor interface model of this type would allow users, regardless of the tactilokinessthetic capabilities of the devices used on the different devices, to familiarize themselves more quickly with the interface of either of these devices. Second, and more fundamentally in relation to the targeted collaborative systems, the main expected result is that this model also allows users to be more effective in their collaborative tasks despite the asymmetry of tactilokinessthetic capabilities offered by their respective devices.
The success criteria of the project will be, on the one hand, the international publication of the results obtained in the best journals or conferences in the field (Revue PRESENCE, CHI, IEEE VR, ACM VRST...), on the other hand, the integration of our solutions with end users. Some manufacturers, external partners of Equipex DIGISCOPE, will be asked to define 3D design scenarios requiring collaborative activities (style study, body design, assembly line design, project modification review...) on which to validate our approaches.
This thesis daily strengthens the collaboration between the VENICE (LIMSI/CNRS) & Ex-Situ (LRI/CNRS & INRIA) team in collaboration with the DIGISCOPE team.
The work itself contributed to the drafting of one of the WPs of a European project submission (Visionair2) involving several teams from this Team, which unfortunately was not selected.