Psychophysiological Interaction and Empathic Cognition for Human-robot Cooperative Work
PsyIntEC (Psychophysiological Interaction and Empathic Cognition for Human-Robot Cooperative Work )is a feasibility demonstration project targeting advances that address safe ergonomic and empathetic adaptation by a robotic system to the needs and characteristics of a human co-worker during collaborative work in a joint human-robot work cell. Inputs to the robot control system will include psychophysiological (or biometric) data from the human co-worker, and this biometric data will provide the basis for affective and cognitive modelling of the human by the robot as a basis for behavioural adaptation.
PsyIntEC addresses the ECHORD Scenario of The human-robot co-worker and hyper-flexible cells and the ECHORD Research Focus Area of Human-robot interfacing and safety, with a secondary emphasis upon Robot hands and complex manipulation in cooperative human-robot tasks. The target industrial scenario is that of a small to medium enterprise (SMI) engaged in rapid prototyping of novel devices. The PsyIntEC project will focus upon demonstrating the feasibility of robotic guidance, support and facilitation of collaborative human-robot prototype production, emphasising support for human emotion and attention regulation, modulation and assessment (e.g. maintaining optimum levels of human attentional engagement in the task at hand) during cooperative human-robot task performance, based upon the use of biometric data to measure human affective, emotional and cognitive states.
Three primary research questions will be explored representing areas of learning effects: RQ1: What are the cognitive and emotional affects on a human worker interacting with a collaborative articulated robotic co-worker during task performance for prototype assembly? RQ2: How can these cognitive and emotional affects reliably be measured in a way that can be used by the robotic co-worker in real time? RQ3: How can knowledge about these states be used by the robotic co-worker to adapt its behavior and/or provide feedback to improve overall human-robot system performance, task objectives and human work satisfaction, health and safety?
Concrete and achievable objectives constituting the proof-of-concept of the proposed advances, relevant to the human-robot co-worker scenario, include: 1. Completion of the construction of a collaborative human-robot work cell as the technology platform upon which the overall experiment will be conducted. The work cell will include a stereoscopic vision system for the robot, two human scale articulated robot arms, an eye gaze tracking system, and psychophysiological interfaces for a human co-worker including EEG, GSR and EMG. The vision system and articulated arms will be organized in an approximately anthropomorphic configuration. 2. Development of a human cognition and affect model (a form of knowledge base) with real-time automated state updates based upon data inputs from biometric interfaces. 3. Development of robot decision and adaptive task performance models informed by the robots human factors, cognition and affect model, used to accomplish empathetic and supportive action generation to achieve reference tasks while optimizing human safety, job satisfaction/engagement and task performance. 4. Development and implementation of a rigorous experimental design and evaluation framework that can be used to design, quantify and compare task performance, decision-making, cognitive and human emotional/affective states of: i) an individual human worker , ii) two collaborating human workers, iii) a collaborative human-robot system in different configurations, in relation to a set of reference tasks.
Contact person: Dr Johan Hagelbäck
Project blog can be found at: www.psyintec.com