Extreme environments are inaccessible to humans owing to factors such as high temperature or high levels of radiation, where conventional sensors and robots cannot be utilized. The Hyper-Environmental Robots Laboratory (HERO Lab) was established to develop robots and sensor technologies capable of replacing humans in such environments. The aim of the HERO Lab is to contribute to work under extreme environments predicted to be required in the future. Undertaken projects focus on inspection of infrastructure facilities, by developing practical products through repeated research and development, field installation, and improvement. HERO Lab has partnered with Tokyo Power Technology Ltd. (TPT) for technology assistance.
Source of photo: Tokyo Electric Power Company Holdings
Japan is known as the robotics capital of the world, where robotics research is thriving; however, during the Fukushima nuclear disaster in 2011, no Japanese-made robot reached the stage of practical use. This is because projects were stuck at the research stage and lacked opportunities and development systems to improve the robots that had been developed by incorporating them into the field. In the future, for several decades, it will be necessary for practical robots to perform tasks in place of humans and assist them in extreme environments such as nuclear decommissioning sites. Accordingly, we need a systematic design and manufacturing system that unifies the maintainability and operability of the various types of robots required for different purposes, make them easy to understand so that they can be operated by workers rather than developers. Overall, it is important to build a track record of robots being utilized at actual work sites and performing useful roles on site. In addition, robots need to work not only with mechatronics but also with complex areas such as sensing and analysis. To bridge this gap, Hakusan Corporation in partnership with the Tokyo Institute of Technology organized a symposium on extreme environment sensing technology in 2019.
At the Tokyo Institute of Technology, Shigeo Hirose conducted pioneering research in robotics and mechanical engineering, contributing to the development of robotics through the invention of several new mechanisms. He is dedicated to establish and implement courses that foster creativity and promote manufacturing education beyond the framework of universities. After leaving the Tokyo Institute of Technology in 2013, he served as the President and Chairman of Hi-Bot, a Tokyo Tech venture. In 2020, he became the director of the Hyper-Environmental Robots Laboratory [HERO Lab], that is developing robots for infrastructure inspection and decommissioning of the Fukushima Daiichi nuclear power plant, thereby inching towards his goal of developing “helpful robots.”
* Recipient of 81 academic achievement-related awards.
＊Individual academic awards received for books and papers are listed in the main publications section below:
Director Shigeo Hirose has been involved in research and development in the following fields:
In 1971, during his post-graduation at the Tokyo Institute of Technology, Shigeo Hirose began researching the propulsive dynamics of living snakes by mathematical analysis and animal experiments. He pursued working in the same field for his doctoral thesis in 1976. This unprecedented biomechanical study of actual snakes was a pioneering step in research on snake-like robots, which has set the standard for biomechanical research worldwide. Till date, no comparable research has been performed anywhere in the world, and the results of this research are still relevant.
In 1972, he succeeded in developing the world's first snake-like robot propelled by the same principles as those of a real snake. Since then, he has led global research on the development of snake-like robots by conducting a wide range of developmental research, including robots designed to explore tight spaces, robots for rescue applications, robots with tactile signal processing, and robots capable of three-dimensional amphibious motion. The serpenoid curve, that is, the curved body shape of a snake when controlling the first snake-like robot, is now used by researchers worldwide as the basic curve for snake-like robots.
His doctoral dissertation and subsequent book on snake-like robots, “Biomechanical Engineering” (Kogyo Chosakai), and its English translation, “Biologically Inspired Robots (Snake-like Locomotor and Manipulator)” (Oxford University Press), is considered as the bible of research for snake-like robots, and have received the following awards:
Hirose also conducted systematic research on snake-like arms, which are long, slender, and can be bent at will. Prototypes have been made of approximately every conceivable arm shape, some of which have been put into practical use. One of these, the “Soft Gripper”—a snake-like robot gripper comprising a small number of drive systems and several joints—achieved soft grasping of arbitrarily shaped objects. This is regarded as pioneering research on inferior drive arms, which has been researched worldwide. Other projects in this area include research on articulated arms driven by coil spring joints with multiple tube wires, research on active endoscopes driven by shape memory alloys, and CT-Arm, which drives each joint from the root with wires and uses “interference drive” to actively apply wire tension to the root joint. He also conducted research on the float arm, which is an articulated arm with a wire-driven weight compensation mechanism.
Among these, the float arm research was put to practical use as an articulated balancer in a joint development project with Nissan Motors, helping to reduce the company’s manufacturing costs, improve quality, and enhance ergonomics. The company received the first Carlos Ghosn award for these results. This was later licensed to CKD, which commercialized it under the name “Power Arm”. This arm was further improved by Hi-Bot Co. Ltd. as an articulated robot arm for plant inspection and is now sold as one of its main products.
Award (1) was for a paper on snake-like robotic grippers, published in 1978. In 2017, it was recognized as one of the 10 most cited papers in the journal Mechanism and Machine Theory since its first publication.
Award (2) was for a paper published in 1988 on a snake-shaped active endoscope using shape memory alloy, which was recognized in 1991 as the most influential paper since 1988 at the International Conference on Robotic and Automation (ICRA), a major international conference on robotics sponsored by the Institute of Electrical and Electronics Engineers (IEEE), the world's largest academic society.
After his research on snakes and snake-like robots, he began parallel research and development on quadrupedal robots, devising a design principle called gravitational decoupled actuation (GDA), an energy-efficient leg design method. In 1981, he was invited to Ohio State University for two months to participate in a project to develop a hexapod walking robot, the adaptive suspension vehicle, which was being conducted by DARPA, a new technology research organization in the U.S. Department of Defense. The proposed leg mechanism based on the GDA principle was adopted for the robot in this project. At that time, the global trend in multi-legged robots was toward six-legged robots. However, Hirose consistently insisted on the superiority of four-legged robots, and since then, global research has focused on four-legged robots.
Subsequently, research on quadrupedal robots led to the development of TITAN 11, a 15-ton robot that functions as a moving scaffold for civil engineering work on slopes. Unfortunately, TITAN 11 has not yet been put to practical use for economic reasons, but this research relays how multi-legged walking robots, which are not yet utilized in industrial applications, could be used in the real world.
In the 1990s, there was a growing demand for research on controlling the gait of quadrupedal robots, but there were no suitable walking robots to conduct experiments. This hindered the progress of walking robot research in Japan. Hirose then developed a quadruped robot as a research platform and made it commercially available at a low cost through a partner company (Okazaki Corporation). This robot has been introduced in laboratories all over Japan.
His research on control methods for quadrupedal robots is summarized in paper ① below, which was awarded the first Best Paper Award by the Robotics Society of Japan. In addition, because it is desirable for robots to be able to run on wheels as well as walk on flat terrain, he also researched “roller walkers” in which the feet turn into roller skates when standing vertically, as well as a robot that can switch between walking and crawling movements.
Hirose believes that it is important to contribute to reducing the workload in dangerous locations where humans cannot work, and is developing humanitarian robots aimed at helping citizens in conflict zones. In 1996, he presented his research on Japan's first landmine detection and removal robot to the Robotics Society of Japan, and participated in the “Research and Development of Sensing, Access and Control Technologies, to Support Detection and Removal of Anti-Personnel Mines from a Humanitarian Perspective” project launched to support reconstruction efforts in Afghanistan by the Japan Science and Technology Agency (JST) in 2002. He visited Cambodia, Croatia, Afghanistan, and other countries to develop a practical landmine-detection robot called Gryphon. He has developed new detection technologies with enhanced detection capabilities using the most trusted and utilized metal detectors in the field, for which he has received an academic awards. The story of development of landmine detection robot is also employed as a learning resource in the second year high school English textbook CROWN (Sanseido) in Japan, under the title “Before Another 20 Minutes Goes By.” Experiments in Croatia indicated that Gryphon can detect landmines more accurately than human landmine detectors.
This research was continued by his co-researcher Fumihiko Edwardo Fukushima, a professor at Tokyo University of Technology, and was employed in a trial mine detection and clearance operation in Angola in 2019.
After developing snake-like and multi-legged walking robots, in the pursuit of more practical forms of mobile robots, Hirose began to explore ways of movement using infinitely rotating devices such as wheels and rollers. An example is a crawler-driven vehicle that can adapt to terrain, for which he has received academic awards ①, ⑤, and ⑦, as mentioned below. This included the development of a new type of lightweight crawler belt and crawler vehicles. The usage of this belt is commercialized by Topy Industries. A few of these vehicles have been utilized for internal investigations at the Fukushima Daiichi nuclear power plant.
Further developments include the development of an “omnidirectional moving vehicle, called VUTON or Vmax. For this work, he received award ② below. In addition, he developed an earthquake simulator with Hakusan Corporation, using this omnidirectional vehicle. This is discussed in detail in Section 6.
He also developed planetary exploration rovers with infinitely rotating wheels for space exploration, for which he received awards ③, ④, and ⑥ below.
Using the principle of the infinitely rotating robot VUTON, in collaboration with Hakusan Corporation, Hirose developed the earthquake zabuton, an omnidirectional vehicle capable of producing any earthquake motion. Previously, earthquake simulators faced issues, such as the need to prepare a huge earthquake generator, operational difficulties in bad weather, and limited width of motion, making it impossible for people to experience a large horizontal shaking earthquake. However, with the development of Earthquake Zabuton, it is now possible to reproduce the same large-amplitude horizontal shaking as an actual earthquake indoors, regardless of the weather. This allows people to experience what an earthquake would be like on a screen and in VR simultaneously and to learn about the relationship between shaking and damage. It is used for disaster preparedness activities in local communities. More than 30,000 people have experienced it so far, and it has been featured in many TV news programs as a new tool for disaster preparedness.
The development of new robots cannot be achieved by simply combining existing elements. Hirose developed several new mechanical elements and sensors to meet these challenges and objectives.
These include a new vision sensor for stereo vision (notable invention in ① below), a magnetic adsorption unit for wall-climbing robots (notable invention in ② below), and an optical 6-axis force sensor (commercialized by Minebea in 2009).
A screw mechanism that can change its pitch in response to load (awards ③ and ④ below), car jack that can be utilized to save lives (award ⑤ below), arm that extends into rubble by pneumatic pressure for rescue (award ⑥ below), rotary load-sensitive stepless transmission (award ⑦ below), several robotic devices for lifesaving (award ⑧ below), grasping mechanism that can efficiently handle disk-shaped objects (award ⑨ below), moving mechanism for a window cleaning robot (award ⑩ below), and a propulsion mechanism for an underwater moving robot (award ⑪ below).
During his time at the Tokyo Institute of Technology, Hirose developed a total of 150 robots and proposed several basic principles of robot design. There have been theories of robot design in the past, but they were all too abstract to be useful in actual design. Hirose’s design principles, developed in the course of the actual design process, are remarkably practical and are still being introduced in lectures at conferences today. Typical examples are the principle of negative energy consumption prevention, connected differential mechanisms, and principle of interference drive. The details are described below.
When Hirose worked at the Tokyo Institute of Technology, he taught design and drafting to third-year undergraduate students in the Department of Mechanical Sciences (now Department of Mechanical and Aerospace Engineering). He launched a course called “Machine Creation” to foster creativity, because simply trying to redesign existing equipment is not design in the true sense. In this course, students created computer-controlled robots to perform funny tricks for an audience (second-year students of the same department), and the grade was determined by how well each robot pleased the audience. This was later adopted as the main competition in the Robot Grand Prix, a robotics competition that he chaired in 1997 to commemorate the 100th anniversary of the Japan Society of Mechanical Engineers, which is now held nationwide.
At the time, there was insufficient space and equipment at the Tokyo Institute of Technology to conduct this kind of manufacturing education; therefore, he sought the cooperation of the university in fundraising and raised approximately 100 million yen to establish an educational space called the “Creative Workshop.” This facility is still in use and shared with the Department of Mechanical Engineering at the Tokyo Institute of Technology.
Hirose believes that it is too late to hone one's engineering sense after starting university education; thus, he holds an annual lecture and robot demonstration competition called “Advanced Robot World” for the general public. He is also the chairperson of the Children’s Challenge Contest sponsored by the Japan Institute for Promoting Invention and Innovation (JIII) and is making efforts for creativity education involving invention clubs across the country.
In Japan, the media has focused on research related to humanoid robots, but Hirose is working on developing robots that are class apart from science-fiction robots. He has highlighted misunderstandings in science fiction-like predictions of the future and ethical issues of anthropomorphic robots, such as Asimov's Three Laws of Robotics, and has continued to propose guidelines for developing robots that are truly useful in people's lives by harnessing their essential characteristics. For these activities, he received awards for the following two articles: