Research
Research
Love of space drives invention for students and researchers
On January 18, 2019, a 100-ton rocket lifted off from Uchinoura in Kagoshima Prefecture. Piercing the crisp blue sky of winter, the 26-meter long Epsilon-4 rocketed heavenward with a payload of satellites developed by private businesses and universities under the auspices of the Japan Aerospace Exploration Agency (JAXA).
These satellites were put into sun-synchronous orbits at an altitude of 500 kilometers. Their main missions were demonstrations of technology needed to enable business in, and utilization of space, a growing arena of activity in recent years. Out of the total 13 themes selected for this launch, Tokyo Tech handled two: an innovative Earth sensor and star tracker applying deep learning (DLAS), and a demonstration for advanced deployable structures based on 3U CubeSats (OrigamiSat-1).
Compared to conventional full-sized satellites, microsatellites can be developed at a lower cost and under a reduced lead-time. Thus they are expected to play a crucial role in space business and utilization. Tokyo Tech is a global pioneer of this satellite format, and we interviewed the head of the Laboratory for Space Systems at the Department of Mechanical Engineering, School of Engineering, Professor Saburo Matunaga. We also introduce the two JAXA-selected projects currently in orbit, as well as two ventures, Axelspace and UMITRON, that are applying Tokyo Tech research to advance private-sector space development.
Tokyo Tech's microsatellite pioneer
Saburo Matunaga
Professor, Laboratory for Space Systems, School of Engineering
Two projects in orbit
DLAS: A deep-learning Earth sensor and star tracker
Yoichi Yatsu
Assistant Professor, School of Science
OrigamiSat-1: 3U CubeSat for demonstrating multi-functional deployable membrane structure
Hiraku Sakamoto
Associate Professor, School of Engineering
University ventures in the space industry
Axelspace: The world's first commercial microsatellite maker
Naoki Miyashita
CTO, Axelspace Corporation
UMITRON: Applying satellite data in aquaculture
Ken Fujiwara
Co-founder/CEO, UMITRON K.K.
We asked Matunaga to talk us through the annals of microsatellite R&D up until the present day, and provide his perspective on the future of space business.
Matunaga: R&D for outer space is enormously expensive and time-consuming; therefore, joint research with national research organizations like JAXA has been the norm up until recently. The microsatellites that we target are like the K Car, a proprietary development made in Japan, if I were to use the auto industry as an analogy. Our goal for this single university lab was to get to implementation. In 2003, the Matunaga Lab at Tokyo Tech and the laboratory of Professor Shinichi Nakasuka at the University of Tokyo developed and launched CubeSats, literally cubes measuring 10 centimeters on each side and weighing approximately 1 kilograms. Since that time, our development has involved nanosatellites, among which CubeSats are the focus.
Although merely a square-shaped box in appearance, the satellite contains sophisticated technology. For instance, in a circular orbit at an altitude of 500 kilometers, the satellite travels on a steady arc at 7.6 km/s, while its status must be confirmed over wireless communications. So a minimum requirement is a communication system and autonomous power supply based on solar cells. A ground station is also necessary for reception of signals transmitted from the satellite and for transmission of commands to the satellite. A mechanism for controlling the satellite's temperature in relation to the ambient temperature in outer space is indispensable, while sophisticated high-performance equipment is required for executing the mission. Technology cultivated on microsatellites gets put to use in large projects with JAXA and related parties. These large projects have massive costs and failure is not an option. Early, multiple demonstrations of leading technologies on microsatellites can reduce risk, which is another benefit.
Matunaga: In 1998, a conference held in Hawaii brought together universities in Japan and the US to discuss new missions for space. At that time, the common perception of a satellite consisted of a mass ranging between 100 and 1,000 kilograms, a development period of five to ten years, and a manufacturing cost of billions to tens of billions of yen. A university lab could not build one alone. But seeing a canned soft drink in front of him, session chair Professor Robert Twiggs at Stanford University suddenly said, "Let's build CanSats and launch them into space." A CanSat is a satellite with dimensions of a soft-drink can. Most of the participants did not take the comment seriously. In fact, only our lab and the Nakasuka Lab at the University of Tokyo engaged seriously on the suggestion and succeeded in developing a model the following year. Our presentations at the next conference prompted Twiggs to suggest the CubeSat. Once again, we were the first to successfully launch a CubeSat.
Tokyo Tech's satellite development history
Matunaga: For better or for worse, we are engineering otaku (nerds). Tokyo Tech has two demonstrations aboard Epsilon-4: an innovative Earth sensor and star tracker applying deep learning (DLAS), and advanced deployable structures based on 3U CubeSats (OrigamiSat-1). The payload planned for Innovative Satellite Technology Demonstration-2 includes a nanosatellite system featuring a variable configuration -- it's a satellite carrying moving paddles and so typical for us engineering otaku. Everyone is engaged eagerly, despite the risk of failure, working on highly complicated technology that ordinarily nobody would want to do. The Tokyo Tech satellites are packed with the latest, leading technologies, you see.
Envisioned motion of a reconfigurable microsatellite
Matunaga: Attitude control is also a highly important function in microsatellites. Mechanisms ordinarily employed are thrusters1 and reaction wheels2, but we newly devised an attitude control method based on multiple moving parts; for example, solar-cell paddles, which normally are stowed during launch and unfold in orbit. Our method controls the orientation of the satellite in orbit by moving these solar-cell paddles. They also can be used for orbital control that exploits atmospheric drag in low earth orbit and solar radiation pressure3 in deep space orbit.
A configurable mechanism, however, tends to become a cause for malfunction, a reason for lack of adoption in the past. Nonetheless, installation of a paddle-drive mechanism allows quick changes to, and maintenance of attitude. We never know when or where the cosmic phenomenon called a gamma-ray burst may occur, for example. During such an event, we would want to immediately alter satellite orientation to face and closely observe the event, but the shift is difficult to accomplish with current microsatellites. We think that our method can aid these observations.
Matunaga: Even if performance suffers in comparison to full-sized satellites, many more microsatellites should be developed and launched in shorter time spans at lower costs. For deep-space exploration, I believe multiple nanosatellites have usage where they can be introduced into asteroid orbits for observation. That's why I am pleased with the launch availability of small rockets manufactured in Japan. For the promotion of business and utilization of outer space in the future, however, I feel the need for a specialized rocket to launch nanosatellites at even lower costs. The roles of a satellite today broadly fall under Earth observation, communications, positioning, celestial observation, or deep-space exploration. The worldwide stance today contemplates accomplishing all of these roles with microsatellites. Eyeing this course, over 100 companies mainly across the US, Europe and China have announced development of a specialized rocket. Although I expect a corporate shakeout to unfold in this arena, I also long for Japanese companies to rise to the occasion.
Matunaga: I wish to engage in the creation of new space business. The center aims to serve as a forum for researchers broadly across disciplines within space science and space engineering to share information and deliberate. The role of a university in particular pertains to the generation of new principles and technology that can serve as a foundation. I look forward to fulfilling that role.
Matunaga: I personally stir the interest and curiosity held by students and support them, but do not offer them any specific themes right away. Research and development of a satellite is a vast effort, but the students are involved because they discovered genuinely interesting facets and significance in that effort. I want them to keep pursuing purely what they want to do, and to challenge the field of space with their particular vision.
Testing 50-kg class satellite TSUBAME
Students preparing to test DLAS
A device that expels gas, liquid, or other material to provide propulsion.
A device that provides torque by increasing or decreasing the revolutions of an internal weight (wheel)
Pressure exerted by light from the sun (electromagnetic waves) reaching the surface of a material. Also called solar pressure.
Yoichi Yatsu, Assistant Professor, School of Science
Overview of the research
We apply a deep learning1 method to recognize images and utilize the technology that identifies land patterns from images taken by satellites. We also verify the operation of a star tracker (an attitude determination device) utilizing commercial off-the-shelf devices in space. This project on the in-orbit recognition of images using deep learning method will be the world's first.
What is the goal of the project?
A satellite is a comprehensive system that includes its ground station, and their telecommunication has become a bottleneck for practical use of small satellites. If we can analyze data and recognize objects in orbit, we can extract and compress data to reduce the latency of data transfer drastically. Furthermore, if the satellites can independently prioritize observations, we can obtain high added-value data more efficiently. In order to drastically increase the accuracy of image recognition, we implemented the deep learning method for experiments on images of the Earth to identify oceans, land masses, vegetation, and land use. In an applied use of the method, we will also conduct experiments to evaluate 3-axis attitude of the satellite by comparing images with maps and actual photos of land masses.
In addition, because my area of specialization is astrophysics, I would like to develop a CubeSat for Astrophysics. Because the light of the stars is faint, the attitude of the satellite must be stabilized relative to reference points to stabilize the camera. The star tracker is used to do this. Although most of the essential components for CubeSats are not manufactured in Japan, I wanted to develop technology that would allow us to manufacture the attitude measurement devices that are fundamental components of these critical parts and essential to advancing the private space industry.
Deep Learning Attitude Sensor control box,
camera unit
What are Tokyo Tech's unique strengths?
DLAS development has been driven by graduate students of Tokyo Tech. This includes the entire range of activity, from concept design to interfacing with JAXA. This shows the great reliability of the graduate students, both technically and personally. Our research team has students who entered Tokyo Tech with the specific dream of working in space exploration, and each has identified exciting areas of research which have led to important contributions to the team. Though recent increases in required coursework have brought challenges to the effort, one of Tokyo Tech's strengths is how it allows students to truly immerse themselves in the research.
Another strength of Tokyo Tech's unique environment is that it gives us the chance to experience failure. We have experienced a wide range of failures in satellite development over the past 15 years. The worst was losing a satellite. In the Japanese manga series, "Space Brothers," Mutta Namba says "Failure is a valuable experience." We strive to use the knowledge we have gained from our failures to advance research and development at DLAS in diverse areas, such as concept design and project management. In terms of facilities, we have benefitted from the Radioisotope Research Center in Ookayama North. Having a radioisotope experiment facility on campus has allowed us to use the most advanced equipment in space exploration. The Radioisotope Research Center has not only benefitted Tokyo Tech, but also manufacturers in the space exploration industry such as Axelspace Corporation as well as universities and research institutions. Unfortunately, however, maintaining this outstanding facility has become increasingly difficult for the university. I hope Tokyo Tech is successful in identifying ways to continue its robust program of industry-academic collaboration to further contribute to the vitalization of the private space industry described by the Cabinet Office's Space Vision 2030
.
Where is the project is now, and where is it going?
DLAS completed in-orbit check-out operations2 on February 12, and the data is now being analyzed. The obtained telemetry shows that the command handling system, two high-performance on-board computers, and six cameras are working without any problems (Figures 2-5). We will use the obtained data to fine tune the parameters and start full-scale experimental observations from late March.
This is an image obtained by processing a monochrome photo. Stars marked with red circles have been successfully identified, and the Hipparcos number is shown for each star. The image shows the direction of Cygnus, with the bottom of the image being north. The Earth is in the upper part of the image and the sun is in the upper right direction.
What are your plans for the future?
I would like to develop an ultra-small astronomical observation satellite. We have worked in cooperation with the NASA Jet Propulsion Laboratory and California Institute of Technology to develop an ultra-wide-field ultraviolet survey observation satellite designed to discover astronomical phenomena such as gravitational wave sources (neutron stars – neutron star mergers) and ultraviolet flash from supernova shock breakout3, and to clarify the physics of these phenomena. We are planning to launch it in to orbit in 2022.
Deep learning is a type of machine learning composed of multilayered neural networks. Studies on deep learning have been conducted since the 1940s. Along with advances in recent computer technologies, deep learning has attracted attention because of its high identification precision for image recognition, etc.
2 In-orbit check-out operations
This is a series of test operations designed to evaluate the functions and performance of devices for satellites. Testing is conducted in the initial phase of operation. After test operations are complete, the initial phase shifts to the stationary phase, in which the satellite is used in missions.
Supernova shock breakout is a flash seen at the moment the shock wave of a supernova penetrates the photosphere of a star. Shock-wave heat instantaneously increases the photosphere temperature to higher than 100,000 K and appears as a soft X-ray or ultraviolet flash. Cooling is rapid, so the flash remains for only about 30 minutes.
Hiraku Sakamoto, Associate Professor, School of Engineering
Overview of the research
A membrane, on which thin-film solar cells and antennas are attached, is compactly stowed into a CubeSat using Origami techniques. We will demonstrate deployment technology for the 1 m x 1 m membrane structure in orbit.
What is the goal of the project?
We developed the satellite to show the feasibility of technology that enables membranes with solar cells and antennas to be folded and deployed, and to obtain data for the design of larger deployable structures. Using five cameras, we will obtain stereo images and videos of membrane deployment to observe deployment behaviors and measure the deployed shapes in orbit. If we can demonstrate the effectiveness in orbit for this structure measurement system that consists of commercially available measurement devices, we will be able to conduct experiments in space more easily. We will also conduct high-speed telecommunications in the ham radio bands to demonstrate downlink methods used to transmit image data to Earth. We will attempt to receive data from space in cooperation with ham radio operators.
What are Tokyo Tech's unique strengths?
A major strength of Tokyo Tech is that it attracts students capable of developing the techniques and devices they need for their projects. New ideas do not occur all of a sudden while discussing issues in front of a white board. Good ideas come after prototypes are developed and parts, systems, or techniques are improved through repeated trial and error. The launch of OrigamiSat-1 was made possible because of the type of student attracted to Tokyo Tech, students capable of developing and implementing ideas in innovative ways.
OrigamiSat-1 unfolding during a ground test at Furuya Lab
Having front-line researchers at Tokyo Tech in a wide range of fields is also an advantage for us. Associate Professor Hiroki Nakanishi
and Associate Professor Hiroshi Furuya
from the School of Engineering, Department of Mechanical Engineering and Specially-Appointed Assistant Professor Takashi Tomura
from the School of Engineering, Department of Electrical and Electronic Engineering participated in the development of OrigamiSat-1. Associate Professor Furuya has conducted research on Origami technology in space structures over the years, and the results of his research contributed greatly to this project. Associate Professor Nakanishi's participation in the planning of a demonstration of the REX-J space robot conducted by astronauts on the International Space Station (ISS) has also contributed greatly. The OrigamISat-1 Project was born of my experience participating in the development of the small solar power sail demonstrator, IKAROS, launched in 2010, and my involvement in the testing of large membrane structure storage and deployment methods. Advancement of the OrigamiSat-1 Project, however, has been driven by the development of a tremendous amount of technology and equipment, including a ground station, and invaluable advice from Professor Saburo Matunaga, who successfully developed the world's first CubeSat.
There is also a system for collaboration with organizations outside the university. For the OrigamiSat-1 Project, Nihon University, Sakase Adtech Co., Ltd., and WEL Research Co., Ltd. participated as members of the development team, and many other universities and companies cooperated in the project.
OrigamiSat-1 (courtesy of JAXA)
Where is the project now, and where is it going?
OrigamiSat-1 was launched at 9:50 am on Friday, January 18, 2019; and one hour and six minutes later, it was successfully released from the launch vehicle at an altitude of 512 km. Four minutes later, the satellite started transmitting radio waves from two deployable antennas. At 11:22 am, an amateur radio operator reported receiving radio signals from the satellite. The Tokyo Tech station also downlinked the radio signals from the satellite at 10:00 pm on the same day, and then confirmed that satellite responded to commands uplinked from the ground station. For the following six days until Thursday, January 24, although there were two unstable operations requiring a reset of the satellite, we were able to obtain data that included satellite voltage and temperature. However, at 8:00 pm on January 24, we experienced a sudden disruption of transmissions from the satellite. We are now trying to identify the cause of the disruption.
What are your plans for the future?
The global trend has been toward the development of electric-powered satellites using electric propulsion and large-capacity communications. Large structures will be required, structures such as solar power generation systems, manned activity stations, and solar panels. We are taking advantage of the OrigamiSat-1 launch to advance research and development of all-electric propulsion satellites for the future. If we can demonstrate the possibility of installing devices on membranes, we will be able to move to the next phase of the development. This will mean expanding structures to a few kilometers in length or the operation of robots on deployed structures.
I believe that maintaining a research environment in which new ideas can be created through the repetition of experiments on the ground and in space for future attempts will not only produce new technology, but will also cultivate the people who will develop new technology in the future. With this belief, I would like to continue my involvement in satellite development.
With space business and the use of space technologies by private companies booming, the 1st Smart Space Device System Symposium held on January 22 at Tokyo Tech's Ookayama Campus came at just the right time. The symposium was hosted by the Research and Development Base for Smart Space Devices and Systems for the Creation of a New Space Industry led by Professor Saburo Matunaga and supported by the Ministry of Education, Culture, Sports, Science and Technology under its Program for the Formation of a Space Collaboration Base. Naoki Miyashita, Director and CTO at Axelspace Corporation, founded in 2008, and Ken Fujiwara, CEO at UMITRON K.K., founded in 2016, both of whom were members of Matunaga Lab at Tokyo Tech, appeared at the symposium to give presentations on the commercial use of small satellites.
Naoki Miyashita, Director and CTO
The company contracted by JAXA to develop and operate the first rapid innovative payload satellite (RAPIS-1) onboard Epsilon-4 is Axelspace Corporation. Founded in 2008, this university-based startup brought together Representative Director and CEO Yuya Nakamura, a University of Tokyo graduate from the laboratory of Professor Shinichi Nakasuka; Director and CTO Naoki Miyashita, a Tokyo Tech graduate from the laboratory of Professor Saburo Matunaga; and other colleagues. Nearly 70 employees represent 13 nationalities and create a highly international setting today.
Business operations range comprehensively from design and manufacture to operation of microsatellites, as well as to provisioning of data analysis for images acquired from satellites. The impetus for founding the business dates back to 2003, when Nakamura, Miyashita, and members during their graduate-school years successfully launched the world's first microsatellites (CubeSats) at their universities. Through a contract with the late Hiroyoshi Ishibashi (founder), Masaya Yamamoto, and others at Japan's private weather information firm Weathernews to develop a microsatellite for monitoring ice floes in the Arctic Ocean, the startup members realized, "Microsatellites have tremendous potential for business in space."
Hinging off of this first contract, the venture initiated a specialized satellite business to develop and operate microsatellites that match client needs. To date, their record stands at five microsatellite launches, including RAPIS-1. WNISAT-1 marked their first launch, the order received from Weathernews for observation of the Arctic Ocean. Their second launch was HODOYOSHI-1, developed under an order received from the Cabinet Office and University of Tokyo. The mission of this microsatellite is remote sensing. With a mass of approximately 60 kilograms, it monitors the Earth from a solar-synchronous orbit at an altitude of 500 kilometers, and can acquire images with a ground-level resolution of 6.7 meters and observational width of 28 kilometers. This track record led to the order received for JAXA's first, rapid innovative payload satellite (RAPIS-1).
RAPIS-1 (courtesy of JAXA)
Miyashita explains, "In terms of resolution, performance cannot match a large satellite. Nonetheless, deploying multiple microsatellites could enable photography of landmass images around the world at greater frequency. Multiple images could reveal changes in nature, crop growth conditions, conditions of a disaster, etc., for the first time."
While advancing in its primary area of designing satellites specialized to customer needs, Axelspace has seen an increase in customers interested in data obtained from satellite image analysis, although not at a level to own a dedicated satellite. In response, the company has initiated its self-financed AxelGlobe business, a satellite-image data platform to launch and operate multiple microsatellites for daily imaging of the entire landmass of the Earth. Through the launch of multiple microsatellites called GRUS with a ground-level resolution of 2.5 meters, AxelGlobe will become a platform that captures data from all points of Earth's landmass daily for database archiving, and then performs analyses through machine learning and other methods. This will enable Axelspace to meet its customers' needs in data and analysis.
Ken Fujiwara, CEO
UMITRON is a venture founded in 2016 by Ken Fujiwara, a former member of Saburo Matunaga's lab. UMITRON accumulates image data from nano & micro satellites and from sensors placed in the ocean. It then analyzes that data to provide aquaculture business operators with useful and timely information.
Along with global population and economic growth, fish consumption has been increasing rapidly. According to surveys, the majority of fishery products in the markets are raised in fish farms rather than harvested from the open seas, and the potential for expanding aquaculture production is approximately 100 times the current consumption of fishery products. However, increasing feed prices have become a significant issue for the aquaculture industry. In addition, the increase of sea water temperature and the outbreak of red tides due to global warming have also become serious problems.
Solving these problems requires the optimization of food supplies and reducing the risk of sudden red tides. Fish are heterothermic. Because of this, they respond to changes in water temperature and their food intake differs accordingly. Therefore, it is important to monitor water temperature to ensure that the optimum amount of feed is provided at the best timing. However, the information obtained from in-water monitoring is limited to small areas. Satellite monitoring would allow us to predict the best timing for feeding according to the changes in seawater temperature over much greater areas, and it would provide more rapid information on the generation of red tides.
Sample marine environment data from JAXA's SHIKISAI climate change observation satellite analyzed by UMITRON
Fujiwara joined JAXA after finishing graduate school. He was involved in research and development of astronomical satellites and probes. He founded UMITRON K.K. with the desire to make space development technology more directly available to society. "After leaving JAXA, I was involved in IT agriculture at a trading company using a satellite remote sensing system. The experience sparked an interest in monitoring marine resources. I wondered if I could apply the technology to the fishing industry. We are conducting currently experiments not only in Japan, but also in Peru, Indonesia and other overseas countries. If the number of nano & micro satellites increases in the future, we'll be able to provide information at more frequent intervals, on the order of 30 to 60 minutes," said Fujiwara.
The Special Topics component of the Tokyo Tech Website shines a spotlight on recent developments in research and education, achievements of its community members, and special events and news from the Institute.
Past features can be viewed in the Special Topics Gallery.
Published: March 2019
© Tokyo Institute of Technology. All rights reserved.
