One of the most important characteristics of human relationships is our ability to gradually get to know and adapt to each other. When we meet a new person, we immediately exchange (via verbal, non-verbal and para-verbal channels) certain crucial information, such as names, approximate age, rough personality traits, etc. This information can then be used in the future for us to adapt the way we interact with a certain person to achieve a certain goal, such as communicating more efficiently or building a prosperous relationship.
The main motto for this work, in general, is to endow social robots with this ability, thus enabling them to build better, long-lasting relationships with human users. Naturally, this is an ambitious goal, towards which we can only take relatively small steps.
In order to achieve this functionality, the following (very high-level) architecture is being developed:
Essentially, the idea is to seamlessly integrate contextual information in the decision-making routines of the robot, infusing each of the robot’s actions with personalization. In other words, the aim is to personalize every action the robot can take, from the execution of GrowMeUp services to low-level functionality such as navigation (through the use of concepts such as proxemics) and prosody.
Naturally, this will require the development of a User Model, which can be used to profile a given user. If enough users can be profiled, we can use the concept of collaborative filtering, popular in the HCI community, to quickly adapt to a new user:
Thus, the robot should be able to adapt both to new and recurrent users.
This work is under development, and has been already been the focus of several publications:
Martins, G. S., Santos, L., & Dias, J. (2015). Towards a Context-Aware Adaptable Services Framework with Applications in Elderly Care. In Workshop on Improving the Quality of Life in the Elderly using Robotic Assistive Technology, International Conference of Social Robotics. Paris.
Martins, G. S., Ferreira, P., Santos, L., & Dias, J., A Context-Aware Adaptability Model for Service Robots, In IJCAI-2016 Workshop on Autonomous Mobile Service Robots, New York.