Social interactive robot navigation based on human intention analysis from face orientation and human path prediction

ROBOMECH Journal, Aug 2015

Robot navigation in a human environment is challenging because human moves according to many factors such as social rules and the way other moves. By introducing a robot to a human environment, many situations are expected such as human want to interact with robot or humans expect robot to avoid collision. Robot navigation modeling have to take these factors into consideration. This paper presents a Social Navigation Model (SNM) as a unified navigation and interaction model that allows a robot to navigate in a human environment and response to human according to human intentions, in particular during a situation where the human encounters a robot and human wants to avoid, unavoid (maintain his/her course), or approach (interact) the robot. The proposed model is developed based on human motion and behavior (especially face orientation and overlapping personal space) analysis in preliminary experiments of human-human interaction. Avoiding, unavoiding, and approaching trajectories of humans are classified based on the face orientation and predicted path on a modified social force model. Our experimental evidence demonstrates that the robot is able to adapt its motion by preserving personal distance from passers-by, and interact with persons who want to interact with the robot with a success rate of 90 %. The simulation results show that robot navigated by proposed method can operate in populated environment and significantly reduced the average of overlapping area of personal space by 33.2 % and reduced average time human needs to arrive the goal by 15.7 % compared to original social force model. This work contributes to the future development of a human-robot socialization environment.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

http://www.robomechjournal.com/content/pdf/s40648-015-0033-z.pdf

Social interactive robot navigation based on human intention analysis from face orientation and human path prediction

Ratsamee et al. ROBOMECH Journal Social interactive robot navigation based on human intention analysis from face orientation and human path prediction Photchara Ratsamee 0 Yasushi Mae 0 Kazuto Kamiyama 0 Mitsuhiro Horade 0 Masaru Kojima 0 Tatsuo Arai 0 0 Graduate School of Engineering Science, Osaka University , 560-8531 Osaka , Japan Robot navigation in a human environment is challenging because human moves according to many factors such as social rules and the way other moves. By introducing a robot to a human environment, many situations are expected such as human want to interact with robot or humans expect robot to avoid collision. Robot navigation modeling have to take these factors into consideration. This paper presents a Social Navigation Model (SNM) as a unified navigation and interaction model that allows a robot to navigate in a human environment and response to human according to human intentions, in particular during a situation where the human encounters a robot and human wants to avoid, unavoid (maintain his/her course), or approach (interact) the robot. The proposed model is developed based on human motion and behavior (especially face orientation and overlapping personal space) analysis in preliminary experiments of human-human interaction. Avoiding, unavoiding, and approaching trajectories of humans are classified based on the face orientation and predicted path on a modified social force model. Our experimental evidence demonstrates that the robot is able to adapt its motion by preserving personal distance from passers-by, and interact with persons who want to interact with the robot with a success rate of 90 %. The simulation results show that robot navigated by proposed method can operate in populated environment and significantly reduced the average of overlapping area of personal space by 33.2 % and reduced average time human needs to arrive the goal by 15.7 % compared to original social force model. This work contributes to the future development of a human-robot socialization environment. Smooth collision avoidance; Social navigation model; Face orientation estimation Background In the recent years, instead of fixed environment like industry, the trend of using robot is shifted to unstructured and public environments such as department store or hospital where humans exist. Same way as computers, many kind of service robots are expected to share and coexist in the same environment as humans to help them with their lives, especially, the kids, elderly and disabled people. Examples of robots’ expected capabilities in public space are avoiding collisions and providing service to humans. Successful methods to achieve safe, time and energy optimized path to avoiding static obstacles or dynamic obstacles [ 1 ] have been achieved for decades. However, to navigate among humans, robot have to come up with a a higher level model because humans adjust their motions related to how the others (human and robot) move and also expect to interact with a robot. This idea is presented in Fig. 1. Three types of general human motion that a robot is expected to be found in a human environment are as follow: • Avoid : humans want to avoid the collision with the robot by themselves. • Unavoid : humans do not avoid the robot and expect the robot to avoid them. • Approach : humans want to interact with the robot. How can we enable a robot to navigate according to a situation where humans want to avoid, unavoid or approach the robot? In this study, we propose a social navigation model (as a unified robot navigation and interaction model in a human environment) which ensure: • Safe and socially acceptable (smooth) motion : a robot does not collide with humans and also do not invade human personal space. • Human priority : a robot avoids the collision in advance in order to provide a free space to human. • Human approach motion : a robot responses to human when he/she approaches robot for interaction. Usually, a robot and a human move in different directions to avoid a collision. In contrast, they get closer when they intend to interact with each other. These behaviors are verified in preliminary experiment and are modeled as a repulsive and attractive force, respectively. We use face orientation to model these forces, as gaze (face orientation) is considered to be a guide of human attention/intention [ 2 ]. Human body pose is considered together with face orientation, to create a modified social force model [ 3, 4 ] for human motion prediction. Human avoiding, unavoiding and approaching trajectories are classified within the range where social space and personal space are concerned. With the proposed model, our robot responds smoothly to human motion. Furthermore, the robot is able to behave like a human by providing the human with face orientation in the intended direction before changing its direction when avoiding collisions, and maintaining a proper distance when it was approached by human. (...truncated)


This is a preview of a remote PDF: http://www.robomechjournal.com/content/pdf/s40648-015-0033-z.pdf

Photchara Ratsamee, Yasushi Mae, Kazuto Kamiyama, Mitsuhiro Horade, Masaru Kojima, Tatsuo Arai. Social interactive robot navigation based on human intention analysis from face orientation and human path prediction, ROBOMECH Journal, 2015, pp. 11, 2, DOI: 10.1186/s40648-015-0033-z