My little artificial companions world – MoCA

One trend in the last twenty years is the increasing number of digital devices used in our daily life; according to prospective studies, this trend will keep accelerating. But be they personal multimedia devices, everyday more sophisticated and integrating more sensors, interconnected, providing intuitive multimodal interfaces such as virtual characters, or domestic robots specialised in monitoring or cleaning tasks, the conclusion is the same: these objects only partially fulfil the users’ expectations. Indeed, they do not yet have the ability to build a relationship with the user, while this ability is considered key to bring significant benefits in daily life. Historically the tool can be considered as an extension of a person. The collection of digital devices must therefore be a true extension of man in his real-world surroundings. He must be able to use this collective, that is constantly evolving, as a single artificial entity embodied in each of the different devices. The artificial companion that arises from this relationship must be reliable, fair, altruistic, safe, fast, credible and empathetic.
The “MoCA” projects aims to advance these devices by endowing them with abilities to build and maintain a long-term relationship with the user. A device will thus be transformed into a true artificial companion.
This fundamental research project will focus on studying artificial companions (virtual characters and personal robots) and their value to users in everyday situations. The artificial companions are interactive intelligent systems designed to maintain a long-term relationship with the user. In order to better understand the concepts of personality, relationship and value, four computing research laboratories (Lab-STICC, LIMSI, LTCI and LIG) with complementary skills and expertise in conducting experiments and psychology will be mobilised.
The ”MoCA” project is original and novel for several reasons: (1)The concept of artificial companions and human-companion relationships will be deepened and enriched, in particular through the notion of personality (2)The communities working on cognitive and affective robots, virtual characters, multi-agent systems and human-computer interfaces will work together synergistically (3)The group of artificial companions includes robots and virtual characters; it is this collective that forms the “Little world of companions” dedicated to a user. Only a connected and interoperable group of companions is able to guarantee permanent access to companionship in nomadic situations (4)The implementation of personalities and behaviours in both the individual devices and in the collective group is technologically challenging (5)The approach is experimental, employing a scenario based on daily-life and with many large-scale experiments. The chosen scenario concerns supporting children, i.e. providing encouragement, emotional support and surveillance, in home situations (6)The societal challenge relates to improving the overall quality of life.
The project will run for 42 months and is organised into 5 large tasks: the first task (T1) scopes the project and gathers the first user requirements; the second task (T2) concerns concepts and models of personality, behaviours and the plasticity of companions; the third task (T3) will develop several demonstrators and a prototype of the « little world »; the fourth task (T4) concentrates on experimentation and the large-scale evaluation of the various demonstrators and prototypes that have been developed. These 4 principal tasks are completed with a task (T0) dedicated to the coordination and the valorisation of the project and its results.
The coordinator of this project is Professor Dominique Duhaut from Lab-STICC laboratory. The funding requested from ANR is 816 525 €.