"Automatic generation of meeting minutes, based on text and graphic annotations, in a distant synchronous collaboration context”.
PHD THESIS SUBJECT, G-SCOP Laboratory, 2014
Phd thesis Title : " Automatic generation of meeting minutes, based on text and graphic annotations, in a distant synchronous collaboration context». Supervisor name : Frédéric Noel Laboratory supervisor : G-SCOP Co-Supervisor name : Philippe Marin Doctoral School: IMEP2 Start Date : septembre octobre 2014 Financing – Context – Partnerships :
Description of the subject :
Collaboration in design teams often requires the involvement of actors from a variety of expertise fields, and thus the use of collaboration tools that allow manipulation of 2D/3D graphic representations of products, and allow keeping trace of decisions and their rationale, often by the way of annotations on these representations. Currently available distant collaboration tools provide little support for this type of activities. Base on the results of recent PhD and master theses conducted in G-SCOP laboratory, this project aims at specifying and developing tools and methods for an “intuitive design office” in which experts may keep concentrated on their technical know-how, forgetting computer tools constrains even is these are always present. The goals of the study is, from a field study in partners companies, to define in details required functionalities, and to develop a computer program mock-up that will be assessed in laboratory design observatory, before being evaluated in actual industrial situations. Objectifs et Travail demandé : This PhD project is set up in the context of distant collaboration for industrial products design, where actors are distributed in various departments of the company, or even from different companies. This technical collaboration context implies, when in face to face meetings, the generation, manipulation, annotation of technical documents that are often graphical supports representing different points of view on the product, related to the domains of expertise of actors involved in the meeting. Building related meeting minutes is often a challenging activity, because of the difficulty of including in these minutes a clear transcription of technical information, drawing or mock-ups elements, that have supported the discussion, argumentation, and are finally the trace of main decisions made (together with their justifications) in the flow of the meeting. Available software tools for distant collaboration (videoconference, application sharing) do not well support synchronous manipulation and collaborative annotation of graphic documents, either 2D or 3D, and do not provide any aid for building meeting minutes. As the results of research conducted for the last few years in G-SCOP laboratory (PhD theses of El-Hadi Mechekour, Onur Isarciklilar, Hanh Vu-Thi, Samira Sadeghi) conceptual bases and software elements are available to demonstrate the feasibility of a distant technical collaboration system around 2D or 3D graphic intermediary objects, integrating annotation facilities and providing a customizable meeting minutes generation. The related research question will study the conditions in which collaborative design could become intuitive enough to allow experts to create or express and share technical contents during project review meetings, take notes easily by the way of annotations in order to keep trace of exchanges, arguments and decision rationale, and finally generate relevant automatic meeting minutes.
Knowledge and skills :
The PhD student will in a first step continue and finalise the development of existing collaboration tools by making them accessible by new interaction technologies available in the lab (multi-touch table, power wall, stereoscopic screen, HMD, immersive 3D desktop,…) to allow single or multi-modal interactions (use of the same environment or collaboration through different environments). From this know-how and on the basis of previous research works, we expect to develop an overall collaboration process that fits industrial needs. The aim of the second step will then be, on the basis of this enrich annotation system, and from the observation of industrial situations, to adapt and finalize the annotation structure and the related meeting minutes generation process. A first mock-up of the final tool will be based on Swhift, then extended to 3D models annotation. Finally, the PhD developments should assess the usability and intuitive aspects of the proposed solution in the context of synchronous collaborative design. A first set of experiments will be conducted and analysed by means of the MEXICO.FR experimental platform. Experimental protocols will be built to objectively characterize the process and the tools. Then usage evaluation should be made in industrial design project context.
The PhD candidate is expected to :
have good skills for analysis and development of conceptual models of design activities and related support tools,
have a minimum experience of technical products design methods, whatever the technical field (mechanics, electronics, …),
have minimal knowledge of computer program development, in order to provide some improvements to existing software tools according to the identified needs,
be able to integrate industrial engineering teams for in situ observation of collaborative design practices.