Notes


A dissertation in HRI: Using Augmented Virtuality To Improve Human-Robot Interfactions

Brief abstract of the dissertation

In this dissertation, in order to improving Situation Awareness, usefulnewss and performance in a variety of navigation and exploration tasks, a prototype 3D interface which presents multiple sets of information in a single correlated window is desinged in comparison to conventional 2D interfaces which typically show different sets of information in different windows.

Some previous work is discussed then the 3D interface was validated by a number of user studies. The results of the studies show that the 3D interface to help operator perform significantly better than a conventional 2D interface.

Finally, three design principles are presented and further work is mentioned.

Summary of each chapter

In chapter 1, the importance of the awareness of the environment around the robot is emphasized. Then the issue of the Poor Situation Awareness issue is identified and the solution is to design a prototype 3D interface which has the two major benefits: a) related sets of information are combined and presented intuitively b) the operator can see more of the environment through a larger field of view.

In chapter 2,previous work in Human-Robot Interactions is discussed and some definitions and concepts are reviewed including some previous work and research results of Interaction Methods and Urban Search and Rescue and furthermore, as well as definition of Situation Awareness and its criteria of Presence, Affordances and Field of View. In Interface Design section, approaches for Conventional Approach, Virtual Environments, Mixed Reality and Augumented Virtuality are discussed and compared.

Chapter 3 focuses on the introduction of the 3D Augmented Virtuality Interface. There are three requirements for a usefull interface, including Information Storage, Integrate Information and Adjustable Display. The technogoy of implementing these three requirements are also discussed in this chapter.

In Chapter 4, the user studies for naviation of a robot through an environment via a conventional 2D interface and the 3D augmented virtuality interface are presented. The results of Path-Following Experiment, Map Building Experiment, Information Usefulness Experiment, Video Size Experiment, Delay Experiment, Real-World Experiment all show positive results that the 3D interface performs much better than the conventional 2D interface.

Furthermore, in Chapter 5, more user studies results in exploration are evaluated, including Pan-Tilt Camera, Find the Foo Experiment. There are some findings in these experiments.

Finally, in Chapter 6, Principles used in the 3D interface design are discussed and validated: a) a common reference frame, b) correlation of action and response, c) and adjustable perspective. Some extensions are also explored, including GPS reference frame, Visualizing camera zoom and Robot arm manipulation.

Overall, in Chapter 7, summary and further work are presented.

Questions

  1. What aspects of the user studies are useful to our project?
  2. How to present the 3D display? Do we have relevant softwares and development environment?
  3. Do we have a developed 3D interface similar to the one in this dissertation?
  4. What model of robots do we have, with arms or not? (Real robot arm manipulation experiment lacked)