Citation Context

... and temporal behavior, and use them to learn a score function representing the preferences in trajectories. User preferences have been studied in the field of human-robot interaction. Sisbot et. al. =-=[34, 33]-=- and Mainprice et. al. [23] planned trajectories satisfying user specified preferences in form of constraints on the distance of robot from user, the visibility of robot and the user arm comfort. Drag...

Citation Context

...its velocity near people [9]. While this method begins to address ideas of planning around people, it does not directly consider social conventions. In contrast, the Human–Aware Motion Planner (HAMP) =-=[10]-=- considers the safety and reliability of the robot’s movement as well as “human comfort,” which attempts to keep the robot in front of people and visible at all times. However, the paths that the plan...

Citation Context

...ters into the personal space of a human carelessly or unwillingly, it has to stop and to wait for a threshold time. Closer to human aware navigation and management of physical space, we could mention =-=[13]-=- where a motion planner is presented which takes explicitly into account its human partners. The authors introduced criteria based both on the control of the distance between the robot and the human, ...

Citation Context

...Path Imitation using a C-Space Constraint For many problems involving human-robot interaction, it is desirable for the paths generated by the robot to conform to some legibility or aesthetic criteria =-=[24]-=- that can be demonstrated via example paths. To incorporate these considerations into our planner we build on a learning-fromdemonstration strategy. We first train a learning algorithm on a set of exa...

Citation Context

...nsidering human’s safety [5], [15]. Besides ensuring safety in robot hardware with compliant designs [20], [1], the motions of the robot need to be planned in a “human-aware” manner. In previous work =-=[17]-=-, [18], we have presented a motion planner that explicitly takes into account human-robot constraints (e.g. their relative distance, the human’s field of view and posture) to synthesize navigation and...

Citation Context

...rd dynamic constraints and planning in such dynamic scenarios. Other work related to planning of manipulation tasks are considering the manipulation tasks embedded in the interaction loop with humans =-=[6]-=-. Considerable research work has also been focused on the development of humanoid robots [7], [8] which are however mostly only equipped with simple and light gripping devices and have relative poor m...

Citation Context

...n of perspective-taking. Considerable attention hassbeen focused on interaction about space, such as direction givings[19, 26], direction interpretation [18, 21], navigation in thespresence of people =-=[28]-=- and walking together [12]. There were ascouple of studies about perspectives for the granularity of roomssize space, such as a technique to learn the names of rooms [29]sand a model ofsthe concept of...

Citation Context

...g behaviors [29].s2.2 Walking with a RobotsSpatial formation has been a focused topic in HRI for the lastsdecade. Social distances with robots were observed and analyzeds[8, 26] and reproduced in HRI =-=[22]-=-. Spatial formation duringsconversations have been replicated in HRI, too, includingsKendon's concept of F-formation [14].sHowever, the above works mainly address static situations wherespeople and ro...

Citation Context

...iant mechanical design of the manipulator [2] and collision reaction strategies [3], preventing collisions in a dynamic and largely unpredictable environment relies on human-phriendly motion planning =-=[4]-=- and, primarily, on the extensive use of exteroceptive sensors [5]. Different types of sensors can be used to detect objects in the environment, including proximity sensors, laser scanners, single and...

Citation Context

... goal positions and our search is uninformed. Gockley [27] observed how older adults walk together. It is reported that partners who were conversing tended to look forward with occasional glances to each other. People navigate obstacles and bottlenecks in a socially acceptable manner by either taking initiative or waiting for the other partner to lead. In [13], they showed that for person following, people have found direction-following behavior more natural than the path-following behavior. Some researchers worked on social navigation to a goal location in presence of people, using proxemics [15, 18, 28, 29]. Althaus [30] developed a reactive behavior to join a group of people and slightly move within the group while interacting. Svenstrup [17] describes a user study where a robot follows random people in an urban transit area. Most people were cooperative and positive towards to the robot. Loper [22] presents a system that is capable of following a person and responding to verbal commands and non-verbal gestures. This work exhibits interaction and demonstrates the robot and the human being in the same team, however it is fundamentally different from ours because there is human intelligence behin...