A demonstration of the force control of the next-generation work equipment Cafe developed in this project was given at a remote construction demonstration by the MLIT.
The Moonshot Research and Development Program
“Realization of Artificial Intelligence (AI) robots that autonomously learn, adapt to their environment, evolve itself in intelligence, and act alongside human beings, by 2050.”
Demonstration of force control using the next-generation work machine Cafe.
Tohoku University conducted a demonstration of ground strength estimation and moving using bucket excavation.
Purpose of sensor pod installation experiment.
We conducted a press release on the outdoor field of Kyushu University for the automated earth and sand transport system.
Demonstration on lunar construction works using small robotic platforms.
We, the CAFE project, are working on the research and development of “Collaborative AI robots for adaptation of diverse environments and innovation of infrastructure construction” in the group of Moonshot Goal #3 “AI robots that learn and act by themselves and coexist with people.” The CAFE project is derived from the Collaborative AI Field robot Everywhere.
Disaster response at the site of
a natural disaster in 2050
In Japan, various natural disasters frequently occur in many areas. To protect the people living in such areas and minimize the damage caused by disasters, techniques for emergency recovery must be developed.
Considering space, the United States and other countries, including Japan, are developing a moonbase under international cooperation, such as in the Artemis Project.
In this project, we have set the moonshot goals of “Natural disaster response” and “Development of a moonbase for manned exploration on the Moon” with a robotic technology that we aim to achieve by 2050.
The left image shows disaster response at the site of a natural disaster in 2050. After accurately analyzing the disaster situation based on sensing data, multiple field robots work together to recover from the disaster by responding to unexpected situations.
Infrastructure construction on the Moon in 2050
The right image shows infrastructure construction on the Moon in 2050. In this illustration, multiple field robots autonomously and dynamically collaborate to construct the infrastructure as well as respond to the unexpected situations.
Initially, the technologies required for disaster response and lunar development may seem different. However, they have one thing in common: they require robotic technologies that can adapt to various situations in various environments. Therefore, our research goal of the 5-year CAFE project is to develop a field robot equipped with such technologies.
In the CAFE project, we aim to design robots that can flexibly respond to various situations. To proceed with this project, we have divided the “ability of responding flexibly” into three elements: “body,” “evaluation,” and “control.” We plan to conduct the research and development in parallel.
To realize a robot “body” that can flexibly adapt to any environment, we will construct “robot hardware that can adapt to any environment.” For flexible “evaluation” within diverse environments, we will research and develop “many modal AI for environmental evaluation,” that evaluate the environment using various sensing data. Furthermore, to achieve multi-robot “control,” we will develop “physical AI and dynamic collaboration” for multiple robots. By combining these three technologies, we can design field robots that flexibly respond to diverse environments. These robots may be used in the fields of “emergency recovery from natural disasters” and “constructing a base on the Moon.”
In the CAFE project, we will promote robotic technologies that can adapt to diverse environments and build infrastructure.
A process diagram of the project is presented here. Although there are many similarities between natural disaster response and lunar infrastructure construction, there are differences in the degree of necessity considering environmental unawareness and robot autonomy.
In this figure, the horizontal axis represents “environmental unawareness,” and the vertical axis represents “necessity of robot autonomy.” For example, the technology required to respond to natural disasters is located at the right side of the diagram because the problem of environmental unawareness is significant compared to the autonomy. However, the technology required for constructing a base on the Moon is located at the upper side of the figure because the necessity of robot autonomy is greater than the problem of environmental unawareness.
Therefore, in the CAFE project, to realize these two technologies, we have named the base technology as “adaptive construction,” and placed it between the technologies of “natural disaster response” and “constructing a base on the Moon.” To realize “adaptive construction,” we will promote research and development in the three areas of “body,” “evaluation,” and “control.”
Simultaneously, we will research and develop more specific robotic technologies required for “constructing a base on the Moon” and “natural disaster response.”
The project named “Collaborative AI robots for adaptation of diverse environments and innovation of infrastructure construction,” aims to develop collaborative field robots that can be useful in “natural disaster response” and “constructing a base on the Moon.” To achieve this goal, robots must be able to flexibly respond to various situations. In this project, we aim to achieve this goal through a variety of approaches.
The introduction to this project is provided in the section “Overview of CAFE Project,” and I, the project manager, would like to share three things that are important to me in this project.
The first is the basic philosophy of this project. The basic philosophy of this project is “to develop technology that is useful to people.” It is the same basic philosophy that I have always cherished in my robotics research. I do not want to excuse myself from saying, “I completed my research, and it will be practical use if somebody complete other required parts.” in this policy.
The second aspect is the horizontal connection between the researchers and engineers. This project requires interdisciplinary cooperation, and it will not be possible without “civil engineering,” “mechanical engineering,” and “AI.” Therefore, we would like to promote the research and development, where researchers from each field respect each other's field and closely exchange information to strengthen the horizontal collaboration.
The third is a field-oriented approach. In this project, we do not want to artificially create and demonstrate a field where robots can efficiently operate. Instead, we want to develop robots that can operate in actual fields.
The members of this project are very powerful researchers and engineers who are currently active in their respective research fields. I am happy to work with them. Please look forward to our future research.
Keiji Nagatani, Project Manager
He received his master’s degree from Osaka University and Ph.D. (Engineering) from the Graduate School of Engineering Science in 1984. He is a professor at the Department of Mechanical Engineering, Graduate School of Engineering, Osaka University, Director of Creative Design Studio on Technology, and Director of the Komatsu MIRAI Construction Equipment Cooperative Research Center, where he researches on robotics and control engineering.
Koichi Osuka
Professor in Graduate School of Engineering, Osaka University
He received his master’s degree from the Graduate School of Tsukuba University in 1993 and Ph.D. (Engineering) from Bielefeld University in 2010. He is currently engaged in the research and development of field robotics for agriculture, civil engineering and construction, as well as marine applications at Yanmar Holdings.
Hisashi Sugiura
Group Leader of Robotics Group, System Research Center, R&D Center, Innovation and Technology Division, Yanmar Holdings Co.
Part-time lecturer at Ritsumeikan and Meijo University
He received his Ph.D. from Yokohama National University, the Doctoral Program in 1990. After working for Toshiba Corporation, Micromachine Center, and Okayama University, he was appointed to his current position in 2014. He is the president of s-muscle Co. Ltd., H-MUSCLE Co. Ltd., and the representative of the new academic field “Science of Soft Robots.”
Koichi Suzumori
Professor of Tokyo Institute of Technology
He received a master‘s degree from Saitama University Graduate School Pre-Doctoral Course in 1995. He received his Ph.D. (Engineering) from Ritsumeikan University in 2014. After working for Hitachi Zosen Corporation and Bomag Japan Corporation, he is currently a senior researcher at the Public Works Research Institute. He is engaged in research on information-based construction and construction robots.
Takeshi Hashimoto
Senior Researcher, Advanced Technology Research Team, Public Works Research Institute
He received his Ph.D. from the Doctoral Program of Engineering, Kyushu University, in 1999. He researches on the applications of geospatial information and disaster prevention based on rock engineering and geo-environmental engineering.
Yasuhiro Mitani
Professor of Disaster Risk Reduction Research Center, Graduate School of Engineering, Kyusyu University
He received his Ph.D. from Tohoku University in 2008. He was a postdoctoral associate at Massachusetts Institute of Technology and a research associate at Japan Aerospace Exploration Agency (JAXA). His research interests include vehicle–terrain interaction mechanics and autonomous mobility for applications in field robotics.
Genya Ishigami
Associate Professor at Keio University
He received his Ph.D. degree from the Graduate University for Advanced Studies (SOKENDAI) in 2011. Currently, he is engaged in space robotics.
Kenji Nagaoka
Associate Professor, Graduate School of Engineering, Kyushu Institute of Technology
He received his Ph.D. from the Kyoto University for Doctor of Science in 1994. He has been engaged in onboard equipment development, systems engineering, and project management for space science missions, and is currently promoting research in the framework of the open innovation hub.
Munetaka Ueno
Chief engineer (research director of future exploration technology), Space Exploration Innovation Hub Center, Japan Aerospace Exploration Agency (JAXA)
He received his Ph.D. from Wayne State University in 2010. He worked as a research associate at Yonsei University, Ehime University, and is currently a project associate professor at the University of Tokyo. His research interest is in civil engineering, especially in the application of ICT technology to the maintenance and management of bridges and tunnels.
Pang-jo Chun
Project Associate Professor, The University of Tokyo
He received his B. Sc., M.Sc., and Ph.D. degrees from the University of Tokyo in 1994, 1996, and 1999, respectively. He is currently a Professor in Computer Vision at Tohoku University, and serves as a leader of the infrastructure management robotics team at RIKEN Center since 2016. His research interests are in the fields of computer vision and machine learning.
Takayuki Okatani
Team Leader of RIKEN, Japan
She received her Ph.D. from the University of Tokyo in 2009. She was a postdoctoral fellow at the Los Alamos National Laboratory, University of Tokyo, an associate professor at Yokohama National University. She is currently an associate professor at the University of Tsukuba in 2019. Her research field is structural engineering.
Mayuko Nishio
Associate Professor, The University of Tsukuba
He graduated from the Faculty of Engineering, Tokai University, in 1983. He engages in the development and research of construction ICT, automation of construction machinery, and unmanned construction technology at Kumagai Gumi Co., Ltd.
Shigeo Kitahara
General Manager, ICT Promotion Office, Civil Engineering Division, Kumagai Gumi Co., Ltd.
He received a master’s degree from the Graduate School of Agriculture, Iwate University Master Course in 1985. Currently, he engages in research on the planning and design of volcano robots for natural disaster prevention during volcanic eruptions.
Toru Shimada
Manager of Business Planning,
Public Consultant Division Kokusai Kogyo Co., Ltd.,
He received his Ph.D. (Engineering) from the University of Tsukuba, Graduate School of Engineering Science in 2002. He is currently engaged in the research and development on intelligent mobile robots, communication technology for robots, disaster response robots, and unmanned construction.
Yasushi Hada
Associate Professor, Mechanical Engineering Program, Kogakuin University
He received M. S. and Dr. Eng. from the University of Tokyo in 1984 and 1989. He became a professor of Research into Artifacts, Center for Engineering (RACE) of the University of Tokyo, a professor at the School of Engineering since 2009, and the director of RACE since 2019. His main research interests are service robotics, distributed autonomous robotic systems, embodied brain science systems, and cognitive ergonomics.
Hajime Asama
Professor, The University of Tokyo
He received his Ph.D. in Information Science from NAIST in 2007. He is currently a tenure-track associate professor and director of the Robot Learning Lab. His research interests include machine learning and its applications in real-world robots.
Takamitsu Matsubara
Associate Professor, Institute for Research Initiatives, Nara Institute of Science and Technology.
He received his Ph.D. from the Tohoku University Doctoral Program in 1986. His main research field is terramechanics.
Hiroshi Takahashi
Professor of Graduate School of Environmental Studies,
Tohoku University
He received his Ph.D. degree from the Department of Mechanical Engineering Science, Tokyo Institute of Technology in 1998. He is a JSME Fellow, SICE Fellow, and RSJ Fellow. His current research interests include legged robot control, computer vision, multiple mobile robots, space robots, service robots, medical imaging, biometrics, and dementia care.
Ryo Kurazume
Professor at the Graduate School of Information Science and Electrical Engineering, Kyushu University
He received his Ph.D. from the University of Tsukuba, Japan, in 2007. After working with the Ibaraki Prefectural Industrial Technology Center, the National Institute of Advanced Industrial Science and Technology, University of Tsukuba. His research interests include robot technology for practical applications.
Toshinobu Takei
Assistant Professor in the Graduate School of Science and Technology of Hirosaki University.
He received his Ph.D. from the University of Tsukuba in 1997. His current research interests include computer vision and computer graphics.
Yasuhiro Mukaigawa
Professor in the Nara Institute of Science and Technology
08/18/2023
Journal
Thannarot Kunlamai, Tatsuro Yamane, Masanori Suganuma, Pang-Jo Chun, Takayuki Okatani (2023), “Improving visual question answering for bridge inspection by pre-training with external data of image–text pairs”, Computer-Aided Civil and Infrastructure Engineering., August, 2023. pp. 1-17
07/02/2023
Journal
Yusuke Tamaishi, Kentaro Fukuda, Kazuto Nakashima, Ryuichi Maeda, Kohei Matsumoto, Ryo Kurazume (2023), “Evaluation of ground stiffness using multiple accelerometers on the ground during compaction by vibratory rollers”, Proceedings of the 40th International Symposium on Automation and Robotics in Construction., July, 2023. pp. 262-269
04/10/2023
Journal
Yuki Kadokawa, Lingwei Zhu, Yoshihisa Tsurumine, Takamitsu Matsubara (2022), “Cyclic Policy Distillation: Sample-Efficient Sim-to-Real Reinforcement Learning with Domain Randomization”, Robotics and Autonomous Systems., April, 2023. Vol. 165, pp. 1-30.
03/22/2023
Journal
Hirotaka Tahara, Hikaru Sasaki, Hanbit Oh, Edgar Anarossi, Takamitsu Matsubara (2023), “Disturbance Injection under Partial Automation: Robust Imitation Learning for Long-horizon Tasks”, IEEE Robotics and Automation Letters.,March, 2023. Vol. 8(5), pp. 2724-2731
03/02/2023
Journal
Huyen Tran, Takayuki Okatani (2023), “Bright as the Sun: In-depth Analysis of Imagination-driven Image Captioning”, Asian Conference on Computer Vision., March, 2023. Vol. 13844, pp. 675–691.
02/15/2023
Journal
Hidenori Takamiya, Ryosuke Yajima, Jun Younes Louhi Kasahara, Ren Komatsu, Keiji Nagatani, Atsushi Yamashita, Hajime Asama (2023), “Reinforcement Learning-based Motion Generation for a Tracked Robot to Go Over a Sphere-shaped Non-fixed Obstacle”, 2023 IEEE/SICE International Symposium on System Integration (SII),. February, 2023. pp. 1-6.
01/25/2023
Journal
Riku Tada, Yusuke Tsunoda, Teruyo Wada, Koichi Osuka (2023), “Proposal of efficient shepherding controller with adjusting gains according to the navigation phase”, AROB-ISBC-SWARM2023,. January, 2023. pp. 1439-1444.
01/25/2023
Journal
Runze Xiao, Yusuke Tsunoda, Koichi Osuka (2023), “The Exploitation of “Unfavorable” Environmental Effects in a Centipede-type Swarm Robot System for Unknown Environment Navigation and Exploration”, AROB-ISBC-SWARM2023,. January, 2023. pp. 879-882.
11/01/2022
Journal
Kang-Jun Liu, Masanori Suganuma, Takayuki Okatani (2022), “Bridging the Gap from Asymmetry Tricks to Decorrelation Principles in Non-contrastive Self-supervised Learning”, Thirty-sixth Conference on Neural Information Processing Systems., November, 2022. vol. 35, pp. 1-12.
10/06/2022
Journal
Tatsuro Yamane, Pang-jo Chun, Riki Honda (2022), “Detecting and Localising Damage Based on Image Recognition and Structure from Motion, and Reflecting it in a 3D Bridge Model”, Structure and Infrastructure Engineering., October, 2022. pp. 1-13.