Proceedings of the 2004 IEEE

International Conference on Robotics and Biomimetics

August 22 - 26, 2004, Shenyang, China

Rescue Robots and Systems in Japan

Fumitoshi MATSUNO Satoshi TADOKORO
University of Electro-Communications Kobe University
Chofugaoka 1-5-1, Chofu, Tokyo 182-8585, Japan 1-1 Rokkodai, Nada, Kobe 657-8501 Japan
International Rescue System Institute International Rescue System Institute
1-2 Minamiwatarida, Kawasaki-ku, Kawasaki, 1-5-2 Minatojima-Minami, Chuo-ku, Kobe, 650-0047 Japan
Kanagawa 210-0855 Japan Email: [email protected]
Email: [email protected]

Abstract— Intelligent rescue systems with high information ”Special Project for Earthquake Disaster Mitigation in
and robot technology have been expected to mitigate disaster Urban Areas” launched by MEXT (Ministry of Education,
damages, especially in Japan after the 1995 Hanshin-Awaji Culture, Sports, Science and Technology) Japan aims at signif-
Earthquake. In development of robots for search and rescue
tasks it is important to develop a robot which can actually icant mitigation the earthquake disaster damage on the scale
work in a real disaster site. Several robots were used for the of the Great Hanshin Earthquake, in big city regions such
search and detection operation in the collapsed World Trade as Tokyo metropolitan area and Keihanshin area. And this
Center building in September 2001. From 2002 ”Special Project research project also aims to establish a basis of science and
for Earthquake Disaster Mitigation in Urban Areas” (5 yeras technologies for earthquake disaster-prevention measures. IRS
project) launched by Ministry of Education, Culture, Sports,
Science and Technology, Japan. It aims at significant mitigation is taking a core role in ”the disaster countermeasure research”,
the earthquake disaster damage on the scale of the Great Hanshin especially Development of Advanced Robots and Information
Earthquake, in big city regions such as Tokyo metropolitan area Systems for Disaster Response in Advanced Disaster Manage-
and Keihanshin area. In this paper, the problem domain of ment System of this project.
earthquake disasters and search-and-rescue processes is clarified
In this paper, the problem domain of earthquake disasters
with introducing an overview of this project.
and search-and-rescue processes is clarified with introducing
I. I NTRODUCTION an overview of this project.

The 2000-year history of Japan in the literatures demon- II. E ARTHQUAKE D ISASTERS
strates that large-scale disasters, such as earthquakes, storms,
floods, wars, etc. have occurred with great frequency. Many Frequency of large-scale earthquakes in Japan had been low
court nobles and feudal lords made special shelters and in the last 50 years until Hanshin Awaji Earthquake in 1995
emergency exits called earthquake shelters and earthquake since Fukui Earthquake in 1948. This period is called a rare
exit in their ancient estates in preparation for sudden shakes. latency period. Recently, large earthquakes have frequently
These names reminds us that Japan has been a world center occurred, and Japan is now in an active period of earthquakes.
of earthquake disasters. The Central Disaster Prevention Committee of Japan Cab-
Intelligent rescue systems with high information and robot inet Office predicts that the magnitude of Nankai Earthquake
technology have been expected to mitigate disaster damages, and Tonankai Earthquake is more than M7-8 [5]. Their prob-
especially in Japan after the 1995 Hanshin Awaji Earth- ability estimated is 40-50% in 30 years. The offing of Miyagi
quake [1], [2]. There are many natural disasters and disasters Prefecture has a 98% probability of more than M7.
by human error, for example accident of nuclear plants. NBC Table 1 shows that Asia has far more large-scale earthquake
terrorism is also serious problem in urban areas. disasters than Europe and America. World Disaster Statistics
In development of robots for search and rescue tasks it indicates that the damage by earthquake is next to that by
is important to develop a robot which can actually work in storms, and is far more serious than that by wars and regional
a real disaster site. Several robots were used for the search incidents [6].
and detection operation in the collapsed World Trade Center The disaster science and technology including seismology,
building in September 2001 [3]. architecture, civil engineering, and geology have contributed
NPO International Rescue System Institute (IRS) has been to mitigate damages by disasters. Recent rather strong earth-
established from April 18, 2002 [4]. It is the industry- quakes did not cause large-scale damage in Japan because the
government-academia-civilian research organization to ad- strength for disaster mitigation has been improved drastically
vance and diffuse high-technologies coping with disaster. IRS by their contributions.
aims to contribute to build a safe society in which people can However, the incidence in Kobe revealed the current disaster
live without anxiety by cooperation of various organizations prevention technologies to be insufficient. Further research and
and human resources. development are necessary by introducing new mythologies.

12
 TABLE I 600
H UMAN DAMAGE BY LARGE- SCALE EARTHQUAKE DISASTERS IN THE

Number of Victims Rescued

21 TH CENTURY
500
Year Country/Region Magnitude Number of Death
1976 China M7.8 242,700
1920 China M8.6 220,700 400
1923 Japan M7.9 142,800
1908 Italy M7.0 110,000
1927 China M7.9 80,700 300
1970 Peru M7.6 66,800 Dead
1935 Pakistan M7.6 60,000
200
1990 Iran M7.3 41,000
1939 Turkey M7.8 32,700
1915 Italy M6.9 32,600
100

Alive
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
III. JAPAN NATIONAL P ROJECT ON R ESCUE ROBOTICS Time after Incident [days]
Japan, which suffered Hanshin Awaji Earthquake, has drawn
Fig. 1. Number of victims rescued by Kobe Fire Department in Hanshin
on the lessons of that experience to alleviate the damage Awaji Earthquake
caused by disasters in major urban areas. One example is the
use of advanced robotics technology in research intended to
create rescue robots. technologies such power assist and mobile mechanisms, elec-
The Central Disaster Prevention Committee has planned the trical technologies such as sensors and mobile communica-
following points as important. Their essences are i) utilization tions, information technologies such as networks and sensor
of civil power and ii) introduction of advanced technologies. fusion, and every technology concerned. The important issue
(1) Practical risk management systems here is to give a solution to this socially important problem.
(1-a) Practical earthquake disaster mitigation and prepared- Every possible countermeasure (not only technologies) must
ness be applied for this final objective.
(1-b) Wide-area response systems Figure 1 shows the number of victims who were rescued
(2) Partnership of citizens for disaster mitigation by Kobe Firefighting Department in Hanshin-Awaji Earth-
(2-a) Local disaster mitigation by collaboration of residents, qauke [7]. The survival rate reduces as time passes. It is well
companies, and non-profit organizations with local govern- Known that rescue in 3 hours is desirable, and that the survival
ments rate becomes drastically low after 72 hours, which is called
(2-b) Collaboration with volunteer works ’golden 72 hours’.
(2-c) Disaster mitigation plan of companies The process of search and rescue in earthquake disasters is
(2-d) Systems for sharing disaster information as follows [1], [7].
1) Aware of existence of victims remaining:
(2-e) Land development of earthquake proof
Voice from rubble piles, information by their family and
(3) Effective efficient disaster mitigation plan
neighborhood, and every other information resource are used.
(3-a) Concentrated countermeasures considering limited
2) Understanding of disaster site situation:
budget
Risk of collapse, gas, distribution of dangerous objects, fire,
(3-b) Earthquake proof of private houses and important
etc. is understood.
public structures 3) Planning of operation:
(3-c) Introduction of economic principles in disaster pre- Procedure, role sharing and logistics are planned. Support-
vention ing parties, machinery and materials are requested.
(4) Full use of advanced technologies 4) Search and Reconnaissance:
(4-a) Advanced information systems Detailed positions of victims and structural damage are
(4-b) Technologies and systems to remove various barriers identified. This information is necessary for human rescue par-
(4-c) Technologies and systems for robustness of modern ties. Relative positions from major landmarks are important,
social systems and absolute coordinates are difficult for real operations under
(4-b) is intended to be development of collapsed structures where base points and lines are missing.
(4-b-i) technologies for information transfer to people who 5) Excavation of rubble piles:
need aids in disasters, Victims must not be injured. The search process sometimes
(4-b-ii) technologies for evacuation guidance, and needs the excavation process, and these two processes must
(4-b-iii) development of robots and systems that work in be discussed together.
inaccessible area. 6) Capture victim bodies:
The word ’robot’ has never used in reports of this field until Victims must not be injured. Flash syndrome should be
then. The field of robots and systems consists of mechanical considered.

13
7) Emergency medicine:
Diagnosis, first aid, care, and psychological support are
performed. Confined space medicine has progressed to give
emergency care in rubble piles in the search process of 4).
8) Transfer victims to hospitals:
Traffic jam and capacity of hospitals are serious problems
in large-scale disasters.
9) Mental care for victims and aiding people:
As symptoms of PTSD are delayed about a month after
traumatic experience, it is hard to make a diagnosis of PTSD
(Post Traumatic Stress Disorder). If symptoms are too delayed
and, appropriate treatment and help could not be received,
stress disorders pass into a chronic state and more serious
situation may occur [8].
Robots and systems are considered to be effective in all over
the process. It is important for this problem to compare new Fig. 2. System overview.
technologies with existing methodologies and conventional
way so that they are evaluated from the viewpoint of users
such as firefighters and disaster managers. How they are better,
under which conditions they are, what constraints exist, how
much do they cost, how many machines are available, when
they become ready etc. are important.
In our national project we have 9 research groups to
accomplish our objective as follows with considering disaster
scenario [9]. Figure 2 shows overview of disaster management
system with disaster simulator and rescue robot system.
• Aero Robots Group (Helicopters, Airships, Balloons):
Main systems of Aero Robots Group are helicopters, air-
ships, balloons. Tasks of this group are global surveillance
(< some km) for information collection at the initial state
of incidents and local surveillance from sky (< 200 m) for
victim search and support of ground vehicles as the second
Fig. 3. Autonomous helicopter (Dr. Nakanishi, Kyoto Univ.)
deployment. Figure 3 shows an autonomous helicopter.
• On-Rubble Robots Group (Crawlers, Wheels, Jumping):
Main robots of On-Rubble Robots Group are crawler type, Tasks of Advanced Tools Group is development of search
wheeled, and jumping robots. Tasks of this group are local and rescue tools for firefighters.
surveillance on the rubble pile (< 50 m) for victim search • Human Interface Group
and environmental check. Figure 4 shows a jumping robot. Tasks of Human Interface Group is development of two-way
• In-Rubble Robots Group (Serpentine, Crawlers, Sensor control interface and data mapping interface. Figure 7 shows
balls): a human interface for robot operation.
Main systems of In-Rubble Robots Group are serpentine • Communication and Data Structures Group
robots, crawler-type robots, and sensor balls. Tasks of this Tasks of Communication and Data Structures Group is
group are local information collection in the rubble pile (< development of communication protocols, ad hoc network,
30 m) for victim search and environmental check. Figures 5 data structures, etc. for data integration. Figure 8 shows a 3D
and 6 show serpentine snake-like robots. mapping by using autonomous helicopter as shown in Fig. 3.
• Underground Robots Group (Wheels, Crawlers): • Field and Evaluation Group
Main robots of Underground Robots Group are wheeled and Tasks of Field and Evaluation Group are development of
crawler-type robots. Tasks of this group are local surveillance fest field, dummy and evaluation method to test developed
in underground structures (< 200 m) for victim search and robots. Figure 9 shows a test field in IRS Kobe Lab. to evaluate
environmental check. developed robots and systems.
• Infrastructure Group (RF ID tags, Micro servers):
Main system of Infrastructure Group are RF ID tags and IV. I NTERNATIONAL R ESCUE S YSTEM I NSTITUTE
Micro servers. Tasks of this group is global information NPO International Rescue System Institute (IRS) has been
collection (> 10 km) using ad hoc networks, micro servers, established from April 18, 2002. It is the industry-government-
RF ID tags, home facilities, etc. academia-civilian research organization to advance and diffuse
• Advanced Tools Group high-technologies coping with disaster. IRS aims to contribute

14
Fig. 4. Jumping robot (Dr. Tsukagoshi, Tokyo Tech) Fig. 7. Human Interface (Dr. Ito, Okayama Univ.)

Fig. 5. Souryu (Prof. Hirose, Tokyo Tech) Fig. 8. 3D mapping (Dr. Nakanishi, Kyoto Univ.)

Fig. 6. Moira (Prof. Osuka, Kobe Univ.) Fig. 9. Test field in IRS Kobe Lab.

15
to build a safe society in which people can live without anxiety announcement of ”Robot Challenge Joint Research and Devel-
by cooperation of various organizations and human resources. opment Project” in the ceremony. Now, under the direction of
IRS has two laboratories in Kobe and Kawasaki. International JFE Urban Development Corporation, another ”hub for rescue
Rescue System Institute (IRS) is taking a core role in ”the robot development” started off. Hereafter, we are striving
disaster countermeasure research”, especially Development to get the importance of rescue activities across to local
of Advanced Robots and Information Systems for Disaster governments and companies, many organizations and people.
Response in Advanced Disaster Management System of this And we are also striving to expand research infrastructure as a
project. base for wide-ranging exchange, transmit information toward
In conformity to the project objectives, we carry out re- the whole world from our laboratory, and become the rescue
search and development of robots, intelligent sensors, portable center of the world.
devices and human interfaces; which could integrate and gauge
the information in disaster confusion and make the best rescue V. I NTRODUCTION OF RESCUE ROBOTS AND SYSTEMS
efforts suited to the situation. We aim to go a long way toward A. Heterogeneous robots in Matsuno Lab.
lifesaving or other human activities and decision-making in The RoboCup Rescue Real Robot League Competition
disaster response by advancing the research and development (Fig. 10) and other rescue robot contests have been held
for active and intelligent data acquisition and integration on worldwide so that we can operate and examine our robots
the network. at simulated disaster sites [10] – [13]. At the competition, we
Advanced disaster response systems against predicted can work out any problems that could occur before we have to
natural/man-made disasters as large as Kobe Earthquake is visit an actual disaster site. We can also polish our technology
now recognized as a necessary advanced social infrastructure. that is required for useful rescue robots by participating in
Research and development of robotics, artificial intelligence, these competitions and experience simulated rescue activities.
computer science, biology, medicine, etc. shall provide inno- We can make use of the results and the experiences obtained
vative solution to this serious problem. Japan is responsible for there for further development in rescue robots.
international leadership of their research and development as
an advanced nation in science and technology. Their applica-
tion and prevalence contribute to establish social infrastructure
for safe and secure human lives. An expanding loop structure
is desirable that these solutions activate related industries and
evolve science, technology and academy, and these activation
and evolution provide new solutions. For the above objectives,
International Rescue System Institute serves research, devel-
opment, prevail, support, international relationship in order to
contribute international evolution and application of advanced
disaster response systems. Its activities shall contribute the safe
secure society and the future science, technology and academy.
As the research organizations taking core rolls in ”Special
Project for Earthquake Disaster Mitigation in Urban Areas”
launched by MEXT, Earthquake Disaster Mitigation Research
Center of National Research Institute for Earth Science and
Disaster Prevention (NIED-EDM) and International Rescue Fig. 10. RoboCup Rescue Japan Open 2004.
System Institute (IRS) have a share in Kawasaki laboratory.
To reach the definite goal of the rescue project, the mitiga-
tion of disaster damage, we should give substance to practical In our multi-robot team we have four different types of
rescue robots and information systems. It is, therefore, fore- robots; tank-type, track-type with two flippers, snake type
most important to communicate with end users. IRS will play and wheel-type with a camera arm. Each type has its own
central roll in the revitalization of the Keihin waterfront area unique control mechanism and possess different advantages
by robot industry. when traveling in various environments.
”Techno-Hub Innovation Kawasaki (THINK)” is the private- 1) MA-1, MA-2, tank-type (Fig. 11, Fig. 12): a remodeled
sector-led project at Watarida in Kawasaki waterfront area. radio control tank which is an off-the-shelf product.
Making heavy use of existing function for supporting research Tracks are driven independently and remotely controlled
and development, this project will establish new business, by an operator using radio signal. For MA-1, there are
support breaking new grounds and realize research activities two CCD cameras mounted on the robot. The first one
on business-academia collaboration in Science Park. And IRS can be driven to pan and tilt with a servo motor attached.
joins the project. The second camera is a panorama camera that provides
On 8 August 2003, ”THINK Mirai-Kobo” for venture firms a panoramic view of the surroundings. The camera
was opened in front of Kawasaki Laboratory, and there was an images are transmitted using a 1.2GHz transmitter to the

16
 operator station. Two front lights can be turned on and
off remotely to assist the operation. For MA-2, we have
installed a camera poll with a “fish-eye” lens camera
attached to the poll. This new configuration of MA-
2 provides the operator a very good view of the field
environment (Fig. 13).

Fig. 12. Tank-type: MA-2.

Fig. 11. Tank-type: MA-1.

2) KOHGA, snake-type (Fig. 14): a custom made robot

developed by our team [14]. The robot consists of eight
individual units. Each unit has two tracks (except first
and last units). These units are serially interconnected
with 2-DOF active joints at both ends first two joints and
3-DOF passive joints in the remaining joints. There are
two cameras each mounted on the head and tail units and
its on-board CPU receives control commands from the Fig. 13. Fish-eye view.
operator via wireless LAN network. The passive joints
contribute good adaptation to complex environments,
and the active joints provide necessary rotation when measure and etc. It is controlled with a Windows-based
certain viewpoint is required. Based on this design interface via wire or wireless LAN network.
structure the snake-type robot has always show high
mobility even in narrow rough terrains. For KOHGA, B. A Novel Teleoperation Method for a Mobile Robot Using
an operator can control the robot by considering only Real Image Data Records
the head unit (rear unit if controlled reversely) by giving In the teleoperation of a mobile robot in a remote site, the
simple command of its movement, as the control method controllability of the robot will increase greatly if the operator
is designed to be a leader-follower type. can easily recognize and understand the situation of the robot
3) FUMA, wheel-type with a camera arm (Fig. 15): The in the remote site and its unknown surroundings.
robot travel around with four wheels (400mm in diame- We have developed a innovative teleoperation method which
ter) and the specially made camera arm not only provide increases the controllability of a robot by using stored im-
surrounding images to the operator but can also be used ages captured by the camera mounted on the mobile robot
to lift up and withstand the weight of the robot. This as spatial-temporal information. This method can deal with
ability comes in extremely useful when stair-climb task the above mentioned difficulties and disadvantages in the
is desired. The robot is equipped with 4 cameras and teleoperation of mobile robots. Plainly speaking, this is the
sensors like heat sensor, laser pointer, ultrasonic distance teleoperation method which uses a bird’s-eye view of a robot

17
 are vaguely known. The proposed method can overcome this
problem and can make full use of the locomotion ability of
the robot and it can also increases its mobility.
This teleoperation method also introduces several other
benefits such as: robustness in low bandwidth communication,
real time synthesis of bird’s-eye view images because of
image-based method and no model construction needed. In
addition, it will also reduce blind spots, prevent the operator
from getting camera motion sickness, and so on.
In this paper we will explain this novel teleoperation method
for a mobile robot using real image data records and different
ways of presenting the synthesized bird’s-eye view images to
an operator.

Synthesis of the bird’s-eye view images to improve remote

Fig. 14. Snake-type: KOHGA. controllability
In our work, synthesis of the bird’s-eye view images,
which improve remote controllability, are conducted using the
following technologies:
• Estimation of the position and orientation of the robot
• Image synthesis technique for bird’s-eye view images
using estimated position and orientation information of
the robot and spatial-temporal information which are
formerly captured real image data records
That is, we need to know the position and orientation of the
robot and its stored real image data records which include
formerly captured images associated with the position and
orientation of the mounted camera where the image was
captured. Overview of the bird’s-eye view synthesis is rep-
resented in Fig. 16. The upper left, center and right pictures
of Fig. 16 are images currently captured by the camera, current
position and orientation information of the robot and the
selected bird’s-eye like image of the robot from real image data
records respectively. The bird’s-eye view of the robot in its
unknown surroundings shown in the bottom picture of Fig. 16
is synthesized using above information and a CG model of the
Fig. 15. Wheel-type: FUMA. robot created in advance.
An operator remotely controls the robot using the composite
bird’s-eye view images which are synthesized according to
in unknown surroundings, is synthesized from spatial-temporal the process presented in Fig. 16 using real image data records
information of formerly captured images. captured by the camera mounted on the robot and position and
This teleoperation method is developed as a rescue robot orientation information of the robot measured by the sensors.
technology. Since it is still difficult to develop autonomous The operator can easily understand the situation of the robot
robots to function well in real environments with current robot and its unknown surroundings in the teleoperation using these
technology, the system structure such as a human operator composite images and the remote controllability will increase.
remotely controlling a robot is one of the realistic solutions
which works well in real disaster sites during rescue robot The algorithm for synthesizing the bird’s-eye view images
operation [14]. Although this method is implemented as part is as follows:
of rescue robot technologies, it can also be applied to any
moving device such as medical surgery support which uses an Algorithm
optical fiber scope. 1) Obtain position and orientation information of the robot
Although high mobility of a mobile robot is indispensable during operation.
for rescue robots in actual disaster sites, it is hard to make 2) Store images associated with position and orientation
full use of the mobility of the robot by ordinary teleoperation information of the mounted camera when they are cap-
methods since the situation of the robot and its surroundings tured to the buffer while the robot is moving.

18
 Fig. 18. Pseudo real-time view.
Fig. 16. Overview of the bird’s-eye view synthesis.

Fig. 19. Teleoperation using a fixed viewing position.

real image data records according to this information of the

robot. Then the model of the robot which is viewed from
the selected image position is superimposed onto the selected
image. The selection of the most appropriate viewing position
is according to the position and orientation information of the
mounted camera which is stored with the captured images in
Fig. 17. System overview. the buffer. As shown in Fig. 18, the selected image is used
as the background image of the bird’s-eye view image. This
background image is not real-time but it is a pseudo real-
3) Select an appropriate image from the stored real image time image. Because of this system configuration it can handle
data records according to the current position and orien- dynamically changing environments in a pseudo real-time
tation information of the robot and make the position and manner. Also this system does not require the construction of a
orientation of the selected image as the viewing position 3D environmental model since this is an image-based method,
of the bird’s-eye view image. and it does not take much time to synthesize the bird’s-eye
4) Render the model of the robot according to the current view image.
position and orientation information of the robot and the
selected viewing position. Teleoperation of a robot using real image data records
5) Superimpose the model of the robot viewed from the In this method the viewing position and the background
selected viewing position onto the selected image from image for the teleoperation is fixed. An operator can see
the stored real image data records (generation of the the robot moving around in its surroundings from the fixed
bird’s-eye view image). viewing position (Fig. 19).
6) Repeat this procedure continuously. One of the examples of this teleoperation using fixed
Overview of this system is shown in Fig. 17. Images viewing position is shown in Fig. 20. The actual images
captured by the mounted camera are stored in the buffer captured by the camera mounted on the robot are shown in
as bitmap images along with the associated position and Fig. 20 (a). The synthesized images using the real image data
orientation information of the camera when they are captured. records and the current position and orientation information of
When the current position and orientation information of the the robot are shown in Fig. 20 (b). The images that are shown
robot is obtained from the sensors, the most appropriate image in the same columns in Fig. 20 are the same time images.
to view the robot in the present time is selected from the stored It can be seen that it is hard to understand and recognize

19
 (a)

(b)

Fig. 20. (a): actual camera images; (b): synthesized images from past camera image data and then the robot position information. Images at each column
in (a) and (b) are images at the same time stamps.

the situation of the robot in its surroundings as well as its R EFERENCES

environment by only using direct images from the mounted [1] S. Tadokoro, T. Takamori, et al., On robotic rescue facilities for disastrous
camera in Fig. 20 (a). It can be said that it is easier to earthquakes -from the Great Hanshin Awaji (Kobe) Earthquake-, Journal
understand the situation of the robot in its surroundings by of Robotics and Mechatronics, 9(1), pp.46–56, 1997.
[2] F. Matsuno, M. Hatayama, et al., GIS and Disaster Simulators in a
using the synthesized images from the proposed method in Prototype System of the RoboCup-Rescue Simulation Project, ICMAS
Fig. 20 (b) rather than the direct images from the mounted Workshop Working Notes, pp.49–58, 2000.
camera. The viewing position of the composite images in this [3] J. Casper, Human-Robot Interactions during the Robot-Assisted Urban
Search and Rescue Response at the World Trade Center, MS Thesis,
example is a fixed point and the image process is conducted Computer Science and Engineering, USF, April 2002.
off-line. The position and orientation information of the robot [4] http://www.rescuesystem.org
is measured by the external sensor. [5] http://www.bousai.go.jp/jishin/chubou/
[6] World Disaster Report, United Nations, 2000
The proposed teleoperation method enables an operator to [7] Tokyo Fire Fighting Department Planning Section, New Fire Fighting
easily understand and recognize the situation of the robot in Strategies, Tokyo Horei Publ., 2002
its surroundings which leads to the improvement of the remote [8] F. Matsuno and T Kawai, Fusion of Engineering, Medicine and Psychol-
ogy in Rescue, Proceedings of the 2003 SICE Annual Conference , pp.
controllability. It can reduce the blind spots by representing a 1102-1103, 2002.
model of the robot as a wire-frame as shown in Fig. 20 (b). [9] Report of MEXT Special Project for Earthquake Disaster Mitigation in
It can also prevent the operator from getting motion sickness Urban Areas, III. Advanced Disaster Mnagement System, 4. Development
for Advanced Robots and Information Systems for Disaster Response,
from the movement of the camera and from also losing sense International Rescue System Institute, National Research Institute for
of direction, especially when the robot is rotating. Earth Science and Disaster Prevention, Japan Ministry of Education,
Sports, Culture, Science and Technology, 2002 and 2003.
VI. C ONCLUSION [10] S. Tadokoro, H. Kitano, et al., The RoboCup-Rescue Project: A Robotic
Approach to the Disaster Mitigation Problem, Proceedings of the 2000
In this paper, the problem domain of earthquake disasters IEEE International conference on Robotics & Automation, pp.4090–4095,
and search-and-rescue processes is clarified with introducing 2000.
an overview of the Special Project for Earthquake Disaster [11] A. Jacoff, E. Messina and J. Evans, A standard test course for urban
search and rescue robots, Proceedings of Performance Metrics for Intel-
Mitigation in Urban Areas. Several rescue robots which are ligent Systems Workshop, 2000.
developed in the project are shown. A novel teleoperation [12] K. Osuka, R. Murphy and A. Schultz, USAR Competitions for physically
method for a mobile robot using real image data records and Situated Robots, IEEE Robotics and Automation Magazine, pp.26–33,
2001.
some robots in Matsuno Lab. are introduced. [13] R. Murphy, J. Casper, M. Micire and J. Hyams, Assessment of the NIST
ACKNOWLEDGMENT Standard Test Bed for Urban Search and Rescue, Proceedings of the 2000
Performance Metrics for Intelligent Systems Workshop, 2000.
This work was done as part of the Special Project for [14] T. Kamegawa, T. Yamasaki, H. Igarashi and F. Matsuno, Development
Earthquake Disaster Mitigation in Urban Areas, supported of The Snake-like Rescue Robot ”KOHGA”, Proceedings of the 2004
IEEE International conference on Robotics and Automation, pp.5081–
by the Ministry of Education, Science, Sport and Culture of 5086, 2004.
Japan. The authors sincerely appreciate contribution of all
members in the project.

20