Game Theoretic Approaches to Advancing Physical Human-Robot Collaborations through Nudging

Create a robotic system that can infer and influence human behaviour to assist humans in physical collaborative tasks. Develop a game theoretic framework to improve human-robot physical collaboration using implicit communication processes, such as intent estimation and nudging as well as explicit feedforward sensory cues. Embed this into an existing methodology of optimal control (OC) based hybrid trajectory optimisation (TO) for realising dyadic collaborative tasks using existing co-bot setup.
Description of the Project: 

Humans perform numerous tasks in their daily life that require collaboration with others. In these scenarios the two humans work together, anticipate how the other will behave and guide their partner towards the joint goal. For instance, two humans can effortlessly move and flip large boxes in a warehouse. Likewise, a human can pour wine into a glass held out by another human, without spilling.

The main focus of the project is to understand the effects of implicit communication in dyadic collaborative scenarios. The goal of the project is to develop a robotic system capable of exploiting intention estimation, game theoretic and nudging techniques to improve the state of the art in physical human robot collaboration. We believe that a successful outcome is an enabler for the broader use of robot partners in the industrial sector, in the hospitality sector as well as in our homes.

The work within this project with focus on the following specific aspects of the system.

(i) Development of an inference module that can predict the human’s intentions and action based on the history of the human’s motion.

(ii) Evaluate sensory feedforward cues that can make collaboration more intuitive and disambiguate agent’s actions

(iii) Utilise nudge theory to influence the human’s behaviour to ease the disambiguation of human intentions.

(iv) Develop a planner based on the ‘theory of mind’ that applies game theory to make decisions and facilitates ‘role taking’ for effective interactions in collaboration.

(v) Exploit the framework above to demonstrate human-robot co-manipulation on existing hardware

Resources required: 
Most large hardware already in place. Consumables for props and several attachments and mock-ups for human robot experiments needed.
Project number: 
First Supervisor: 
University of Edinburgh
First supervisor university: 
University of Edinburgh
Essential skills and knowledge: 
Reinforcement learning, control theory, nudge theory, role-taking, game theory, optimisation
Desirable skills and knowledge: 
ROS, hardware experience, real-time control, floating base robot platforms
Funding Available: 
  2. Li, Y., Carboni, G., Gonzalez, F. et al. Differential game theory for versatile physical human–robot interaction. Nat Mach Intell 1, 36–43 (2019).
  3. Theodorous Stouraitis, Iordanis Chatzinikolaidis, Michael Gienger and Sethu Vijayakumar, Online Hybrid Motion Planning for Dyadic collaborative Manipulation via Bilevel Optimization, IEEE Transactions on Robotics (T-RO) (2020).