11:00-11:30
"Surgical Robots in LESS and NOTES"
Dong-Soo Kwon : Department of Mechanical Engineering, KAIST, Korea
Asbtract
Incision of abdominal wall, pain and infection are the problems of current open-surgery and laparoscopic surgery. To overcome these problems, LESS (Laparo-Endoscopic Single-Site Surgery) and NOTES (Natural Orifice Transluminal Endoscopic Surgery) were proposed. LESS inserts surgical tools through umbilical incision and NOTES inserts surgical tools though natural orifices, which reduces pain and complications compared to the open-surgery. However, using current platforms, LESS and NOTES holds greater limitations that do not exist in open-surgery or laparoscopic surgery. Thus, utilizing robotic technology to overcome these limitations is in progress. This paper introduces researches on LESS and NOTES and delineates requirements.
11:30-12:00
"Human Robot Interaction Framework for Conscious Robots"
Jaeyeon Lee: Electronics and Telecommunication Research Institute (ETRI), Korea
Asbtract
As an intelligent entity, robots are supposed to be conscious of the environments. Especially in the perspective of HRI, robots should be aware of the humans around them, not only their existence but also their identities, their whereabouts and their behaviors. Further the robots are expected to do so without explicit cooperation of the human part. Many sophisticated algorithms including human detection, face recognition, speech recognition, human tracking, behavior recognition and sound source localization have been developed and applied to make the robots intelligent. However, through experiments, we found that each recognition algorithm alone cannot achieve the real consciousness for the robots. The algorithms should be applied at an appropriate time in an appropriate order. Also, the algorithms should be always up and running rather than being called when the application part needs them, which of course causes serious resource problem. The HRI framework is software that choreographs various HRI components in an efficient way. Also, the HRI framework provides a standardized interface with the application programs. In this presentation, we want to discuss the huddles in developing conscious robots and introduce the framework as a way to overcome such huddles.
12:00-12:30
" Dexterity of Underwater Robots"
Sadao Kawamura : Department of Robotics, Ritsumeikan University, Japan
Asbtract
Oceans are being used more intensively for a multitude of uses including the development of energy sources, mining of rare metals, and fish farming. Maintaining sufficient supplies of clean and drinkable freshwater is one of the most important and imminent tasks. Robots are expected to play an important role in carrying out tasks underwater in both ocean development and freshwater preservation. In general, the maneuverability of underwater robots has been researched on AUVs (Autonomous Underwater Vehicles) or ROVs (Remotely Operated Vehicles). In contrast, the ability of manipulation of arms installed on underwater robots has not been intensively studied.
A new underwater robot with arms is presented in this talk. The robot has dual arms and movable floating blocks to stabilize the robot in order to dexterously perform underwater tasks. A project conducted on environmental problems of Lake Biwa with this robot is introduced, and some experimental results are presented from this and other projects.
12:30-13:00
" Temporal Bayesian Network for Proactive Assembly-Assistant Robot"
Il Hong Suh : College of Information and Communications, Hanyang University, Korea
Asbtract
Timely interaction between a robot and a human is an important factor for an assistant robot in human-robot coexistence environment. For such interaction, a robot has to predict future situation related to human, and has to proactively execute behaviors to be appropriate in the predicted future situation. In this work, we will present a novel temporal Bayesian network for probabilistic temporal prediction and proactive action selection in the continuous time domain. Using our proposed network, both temporal and causal information can be inferred simultaneously in a unified framework. Then, as an application example, we will introduce a proactive robot system to assist a manual assembly task, where a robotic assistant prepares in advance a necessary part for the next step of assembly, even though a human does not ask the robot to get the part for the assembly.
13:00 - 14:20 : Break
14:20-14:50
"Intelligent Software Modules for Interactive Service Robots"
Tsukasa Ogasawara : Graduate School of Information Science, Nara Institute of Science and Technology, Japan
Asbtract
Demands on interactive service robots for human-robot symbiosis are increasing. Considerable attention has been given to research on communications between humans and robots. Interactive functions and system integration are key technologies for such robots, and researches have focused on various technologies those make robots actually useful in our daily life. We have been focusing on research on real time sensing to understand human behavior, such as real time gaze measurement, multi-modal communication, and vision-based interfaces. In this talk, these human modeling technologies are introduced. Furthermore, interactive service robots and human-robot interaction modules for such robots are presented.
14:50-15:20
"Development of a wearable sensory system for walking-support exoskeletons"
Young-Jo Cho : Robot/Cognitive System Research Department, Electronics and Telecommunications Research Institute(ETRI), Korea
Asbtract
Walking-support exoskeletons for disabled are typically controlled by a sensory system reading the user?fs intention of walks. We have been developing a wearable sensory system including FSRs(Force Sensing Resisters) and IMUs(Inertial Measurement Units), which recognizes the walking intention appearing while a person is walking with an exoskeleton and Lofstrand crutches. FSRs inside gloves are used to detect the intention of starting a stride, and those inside shoe-soles give clues to the walking phase identification. An IMU enclosed in the sensor shirt reflects the pose of the torso. The extracted intention is finally used to control a walking-support exoskeleton assisting the lower-limb disabled.