An astrometric satellite calledNano-JASMINE will be launchedin 2013

 from the Alcantara Launch Center in Brazil, with the aim of precisely mapping the position of approximately 200,000 stars across the sky. Developed as part of a joint project between the National Astronomical Observatory of Japan (NAOJ), the University of Tokyo (Todai), Shinshu University and Kyoto University,Nano-JASMINE is a nanosatellite weighing just 35 kg. In spite of its small size however, there is nothing modest about its mission. It is designed to precisely map the three-dimensional position of approximately one million stars across the sky. Pinpointing the position of stars will help us to understand the distribution and movement of dark matter, which in turn should hopefully provide an insight into the structure and history of space itself.

Twenty-two years ago, in 1989, the European Space Agency (ESA) launched a 1.4-ton satellite calledHipparcos with the same aim. The enormous challenge currently being undertaken byNano-JASMINE is to map the stars with roughly the same degree of accuracy but using a mini-satellite that weighs 1/40^th and cost less than 1/100^th as much as its predecessor.

Whereas the on-board measuring instruments were developed by NAOJ, the satellite itself was built by Todai. We may have the ability to produce state-of-the-art scientific satellites like this now, but we only started developing satellites a mere ten years ago. The first satellite we developed was a tiny cubic satellite that measured 10cm and weighed 1kg. It was a CubeSat satellite designed in accordance with international nanosatellite specifications. CubeSatXI-IV (sai four) was launched from the Plesetsk Cosmodrome in Russia on June 30, 2003, and went into orbit around the earth at an altitude of 824km. TheXI-IV marked the starting point of our voyage into satellite development.

Satellites on a shoestring

CubeSatdevelopment started as a result of a model satellite project called CanSat, which became CubeSat’s predecessor. At a US-Japan academic conference in Hawaii in 1998, attended by students and teaching staff specializing in space engineering, Professor Bob Twiggs from Stanford University held up a coke can and proposed “let’s make this into a satellite” (Bob Twiggs,CubeSat: Satellites for Everyone, p. 32). The aim was to install communication devices, attitude control systems and other equipment inside cans, launch them several kilometers into the sky on board rockets made by amateur groups in the United States, and then bring them back down to earth by parachute, conducting a range of experiments on the way.

The first CanSat event took place the following year, in September 1999. Four academic institutions took part, namely Todai, Tokyo Institute of Technology (Tokyo Tech), Arizona State University and Redwood High School. Amateur rockets were launched out in the desert in Arizona and satellites released at an altitude of 4km.

At that year’s conference in Hawaii, which took place soon after, Professor Twiggs stood up again and showed everyone a 10cm square plastic box. “This year, let’s make one of these into a satellite.” Rather than launching a satellite from the ground and letting it fall back down to earth however, the aim this time around was to produce an actual satellite that would go into orbit around the earth.

The size and weight of each satellite would be standardized, to enable the same mechanism to be used to release satellites from the rocket, and to simplify the design process. That would mean that satellites made by individual teams could be brought together and launched at the same time, and would also make it easier to negotiate arrangements with launch providers. The ability to share component and design-related expertise would be an added bonus too. As they listened, the audience became increasingly excited. Twiggs had effectively sparked off a race to develop CubeSat satellites.

Having already been involved in the CanSat project, Todai and Tokyo Tech got straight down to development. As our first attempt at developing a satellite, this meant learning everything from scratch along the way. Our top priority was to cut out any unnecessary systems and create the simplest possible structure. We fitted a permanent magnet and decided to use a passive attitude control system that would be geomagnetically guided in the right direction. We opted not to use an active attitude control system, which would have used sensors to detect and adjust the direction of the satellite, on the grounds that it would be too complicated.

We weren’t over ambitious with the mission. As long as we could test the communication devices, attitude control and other basic satellite technologies, that was fine. There were countless items that we wanted to add, but it was essential that we knew our limitations. By way of an advanced mission, albeit something of a long shot, we decided to install a mini-camera in order to take images of the earth.

It required some precise maneuvering to fit all of that equipment inside a 10 cm cube. We thought long and hard about how to arrange everything, so that the four substrates containing the electronic components fit into the gaps without touching one another. In the absence of wings, the solar cells used to power the satellite were attached to the outer surface. The small size of the satellite meant that the surface area was limited, resulting in an average generating capacity of 1.1 watts. That had to be enough to power everything, including communications.

We used a processor with exceptionally low power consumption for the continually operating on-board computer, turning a blind eye to performance to some extent. We similarly asked Kobe-based wireless equipment manufacturer Nishi Musen Kenkyusyo Co., Ltd. to provide us with a low-power communication device which fits into 10 cm cubic space. Incidentally, the communication device in question went on to become a best seller, and is now used in the majority of Japanese nanosatellites.

The fact that we had very little money made it even more of a struggle. For the solar cells, we used general-purpose monocrystalline silicon cells rather than gallium arsenide cells designed for use in space. We bought most of the components from shops in Akihabara or online. Students would write down items that we needed on a blackboard in the lab, and then take it in turns to copy down the list and cycle over to Akihabara every day. We were thankful that the university is in such a good location, close enough to cycle to Akihabara.

We decided to test our newly developed communication device in an environment that was as realistic as possible. With the help of the ISAS Sanriku Balloon Center (Ofunato, Iwate Prefecture), we attached the communication device to a balloon and sent it up to an altitude of 40km. We then conducted a communication test lasting roughly an hour, between the device and an earth station approximately 500km away back at Todai. We were also able to simulate satellite operations, including sending instructions to the satellite and downlinking data from the satellite (sending data back down to earth), taking us a step closer to launch-readiness. As we obviously couldn’t set up testing equipment in advance, we went around borrowing equipment from other facilities. We conducted vacuum experiments at another Todai laboratory, and radiation tests at the Takasaki laboratories belonging to the Japan Atomic Energy Research Institute (JAERI, now the Japan Atomic Energy Agency (JAEA)) and facilities owned by the National Space Development Agency of Japan (NASDA, now part of the Japan Aerospace Exploration Agency (JAXA)). We conducted thermal vacuum tests and vibration experiments meanwhile at the Institute of Space and Aeronautical Science (ISAS, now also part of JAXA), and radio wave testing for the communication device at an anechoic chamber belonging to the Tokyo Metropolitan College of Aeronautical Engineering.

We saved money on radiation testing by getting another organization to position our satellite next to the test bench whilst conducting its own radiation tests. We put the satellite inside a small duralumin case, which students then carried by hand. It was fortunate that the satellite was small enough to enable us to do that. By the end, we had managed to carry out almost all of the environmental tests required for a regular satellite.

No launch?!

We really began to have difficulties when it came to finding an opportunity to launch the satellite. Nanosatellites have to be launched using what is known as a “piggyback” approach, which involves attaching them to the side of a full-size satellite and loading it onto a rocket. As JAXA had no such plans for the time being, we had no option but to look to other countries.

The year after we started our research, a company called One Stop Service to Space (OSSS) approached us at the conference in Hawaii. OSSS was a venture company set up by a group of US university professors to provide launch opportunities on board the Russian rocketDnepr. We were only too happy to take them up on their offer and paid them a contract fee of ¥3 million.

Despite providing information on our satellite as instructed by OSSS and their agent in Japan, we received no notification of a preliminary design review (PDR) meeting, which would usually be essential at the launch preparation stages, and had absolutely no idea when the launch would go ahead. In early 2002, a foreign student from Bangladesh who happened to have been assigned to my lab mentioned that he had previously been based in the Ukraine. We therefore got him to contact ISC Kosmotras, the company responsible for launchingDnepr, via a friend and find out what was going on.

Our inquiries yielded some surprising information. ISC Kosmotras said that it had never entered into a contract with OSSS to provide launch opportunities. It turned out that OSSS had failed to secure the amount of money it had expected, and had been unable to sign a launch contract. Our money had been taken for nothing. It was my own fault for failing to check the contract properly. I really should have made sure that there was a clause in the contract stating that we would get a full refund if the launch failed to go ahead. That was a major blow and an expensive lesson, especially as we had such a limited budget.

We later heard that there would be a launch opportunity in the United States. We took our satellite over to the United States only for the launch to be cancelled at the last minute. We were really struggling to secure a launch, and knew that if we delayed it much longer the students who had worked on developing the satellite would graduate. In the meantime, we completed work on theXI-IV in the fall of 2002.

We no longer had time to spare. As a last resort, we wrote a letter saying “we only have ¥2.5 million, but please would you launch our 1kg satellite?” and sent it to Eurockot Launch Services, a launch company established as part of a joint venture between Germany and Russia. To our surprise, we received a reply asking us to come to a rocket PDR meeting in Moscow in December. It said that, if we could agree the details, we would be able to put our satellite on the rocket. We jumped into action and started preparations in partnership with Tokyo Tech, which had been developing a CubeSat at the same time. In December, I headed over to Moscow on my own, braving -20˚C temperatures.

Matsunaga Saburo, the head of the team from Tokyo Tech, was supposed to come with me, but I received a fax at my hotel that night saying that he would be unable to come due to visa problem. The following day, the rocket PDR meeting took place at the Proton Hotel, opposite the company making the rocket, Khrunichev. Despite feeling like I had entered enemy territory, I explained our plans as best I could. They agreed that the satellites would be of great educational value and would pose no technical issues in terms of fitting them to the rocket, and decided there and then to launch both the Todai and Tokyo Tech CubeSat. I was amazed and extremely grateful to the Russians for the speed of their decision and their tremendous generosity.

Things moved quickly after that. We signed a launch contract the following year, in January 2003, and took our satellite over to Russia in a suitcase, along with a group of seven students, in June. The Tokyo Tech team came with us. We were buffeted along on a sleeper train from Moscow for 18 hours, to the launch base at the Plesetsk Cosmodrome. Once we had undergone a final check at the launch base, which was essentially a secure town where you couldn’t even take photos, we were able to assist engineers from Eurockot at every stage after that, right through to fitting the satellites to the rocket. It was an amazing experience to catch a glimpse of Russian space exploration first hand like that.

SendingXI-IV into space

The entire team returned to Japan after that. As the launch date approached however, we sent two students over to the control room in Russia and got ready for the big moment. We were hoping to watch the launch from Japan, via images relayed by satellite byEurockot, but that turned out to be unfeasible in the end. It had cost ¥3 million to develop the satellite and ¥2.5 million to launch it, but it would have cost ¥5 million to broadcast images of the launch. Ultimately, we had to make do with online updates sent from the students at the launch site.

The date of the launch arrived. At 23:15 (Japan time) on June 30, the rocket carrying satelliteXI-IV finally took off. Our satellite was scheduled to detach from the rocket at 00:46. The students all came to the university lab and waited for the first radio waves sent fromXI-IV as it passed over the skies above Japan.

At 4:36 the phone in the lab rang. I picked up and heard the words “Got it!” It was one of our students, who had been assigned to the Sugadaira Space Radio Observatory thanks to assistance from the University of Electro-Communications (UEC). Shortly after that, the satellite passed over the skies above Tokyo. In amongst all the noise produced by the receiver, we could hear the beacon signal from theXI-IV, quiet but unmistakable. We all joined hands without thinking. Some of the students even shed a tear or two. Then we toasted our success. After years of hard work, we had got our reward and wanted to celebrate that glorious moment. The first we heard was an email from an overseas amateur radio operator, who claimed to have picked up radio waves from the Tokyo Tech satellite, which had been launched together with our own. In actual fact, both Todai and Tokyo Tech were using amateur radio bands to communicate with their satellites. As we had published details of the frequencies, amateur radio operators all over the world were listening out for us. We continued to receive reports of radio waves from the Tokyo Tech satellite being picked up, but there was no news of the Todai satellite. We were understandably starting to get frantic.

The first task in terms of operating the satellite was to calculate its orbit. Shortly after the satellite was released, eight sets of orbit data were published by North America Aerospace Defense Command (NORAD), which monitors space objects. For several days after that, debate raged between amateur radio operators around the world as they tried to work out which orbit belonged to which satellite. We were obviously trying to calculate the satellite’s obit too, based on radio waves reaching earth stations and Doppler shift. Although we couldn’t find an orbit whose figures matched our projections exactly, we ascertained that this was down to a slight change in frequency, due to factors such as the temperature of the oscillator on the transmitter. Within a few days, we had identified the orbit of theXI-IV.

Despite the harsh conditions, bombarded with radiation within a vacuum and subject to massive variations in temperature, all of the equipment still worked normally. The on-board computer gave us cause for concern as it reset itself nine times during the two months after launch, but we worked out that this was due to a minor bug in the software rather than radiation, as we had feared. As it wasn’t particularly frequent, it had virtually no impact on the satellite’s operations. By about two months later, we had completed all of our primary missions, including testing communications.

 All that remained was our “long shot” mission to take images of Earth. As we weren’t controlling the direction of the satellite, we could only take images when it happened to be facing towards Earth. Given the slow downlink speed of just 1,200 bits per second, taking random images and sending them back to Earth was also out of the question. We decided to take seven images at different times and attach headers containing details such as the percentage of dark space and average brightness. We started by downlinking the header data and then prioritized images that were most likely to be good. We downlinked low quality images first of all. If any images looked promising, then we moved on to more detailed images. That was how we managed to get hold of images of Earth taken from space.

We organized an email service to send copies of the images to anyone interested, in an effort to familiarize members of the public with CubeSat. The number of subscribers has increased with every satellite launched since then, with the registered total currently standing at over 3,000. The reaction has been even better than expected, with some registered subscribers even sending in messages of support. Encouraging emails from female subscribers in particular seemed to make our students really happy. What space exploration is lacking at the moment is a way to feed results back to the general public in way that anyone can understand. I feel that this trial sets out an example of one way in which we could achieve that.

Students operated the satellite on a rota basis on alternate days. The satellite crossed the skies above Japan at roughly the same times every day. It averaged out at three times a day – once or twice around 4:00 am and once or twice at around 4:00 pm – making for an overall operating time of approximately 40 minutes. To begin with, four or five people would get involved in operations every time, but that number gradually tailed off. By around a year and a half later, the satellite was being operated by one person at a time.

Working with a large number of amateur radio operators, we also set up a system whereby they could forward any data they received from the satellite to the station at Todai via email. Before long, we developed software so that downlinked data was forwarded automatically to the Todai station and got users to install it on their radios. This meant that, whenever a downlink order was sent from Todai to the satellite, the data from the satellite was then received by amateur radio operators all over the world and sent back to Todai on a real time basis. So even if we couldn’t receive data at Todai, there would be no need to downlink the data again as long as it could be received somewhere else in Japan. As there is so much noise in Tokyo, it is often impossible to decipher radio waves even once they have been received. That’s when we draw on help from amateur radio operators in outlying areas, which has substantially improved the efficiency of our operations.

As the number of satellites launched by our lab increased after that point however, we began to struggle to handle all of them ourselves. Since 2008, we have left theXI-IV entirely in the hands of amateur radio operators. It is still fully functioning and taking images of Earth to this day, more than eight years after being launched.

Sending a second satellite into space

We made a backup version ofXI-IV in advance. It was the same as the original satellite and was intended as a replacement if there were any issues immediately prior to the launch, or as a means of identifying the causes of any malfunctions in orbit. It was the backup version that enabled us to run tests and trace the mysterious computer resetting issue to a software bug.

As theXI-IV continued to fly according to plan however, the need for a backup became less and less. We started to look for an opportunity to launch the backup satellite too, and found one from a somewhat unexpected source. The University of Toronto Space Flight Laboratory (SFL), which had launched a CubeSat at the same time as theXI-IV in 2003, had secured a launch opportunity on board the Russian rocketKosmos. As SFL wasn’t going to finish developing its satellite in time however, it was willing to hand over its launch opportunity to Todai, for a fee.

SFL had been given the launch slot in question by the ESA, which was sending up an educational satellite built jointly by European students calledSSETI Express, in return for developing a release mechanism that would release multiple CubeSat. Once in orbit, the rocket would releaseSSETI Express, which would then release the CubeSat. In fact, the release mechanism was something that we had designed ourselves, and then provided the University of Toronto with our drawings as part of a joint research project. By a curious coincidence, our satellite was going to be carried by the release mechanism that we had designed.

This meant a four-way contract between the ESA, as the developer of the main satellite, Kosmos, as the company launching the rocket, SFL, as the recipient of the launch slot, and Todai. Contract negotiations were complicated and slow going. Fortunately however, the Advanced Program Manager at SFL was Freddy Pranajaya, an Indonesian friend of mine from my time at Stanford University. He acted as intermediary with the other three parties in the contract negotiations, so I felt deeply indebted to him. As a result of all this, it was decided that our backup satellite,theXI-V would be launched along with other CubeSat from Norway and Germany in summer of 2005.

As we were going to the trouble of launching another satellite, we wanted to give it a new mission. Just as we were thinking about the nature of that mission, we heard from Kawakita Shiro, from the Space Power Systems Group at JAXA (Aerospace Research and Development Directorate), that JAXA had developed a new type of solar cell that was more resistant to radiation and was looking for an opportunity to test it in space. We therefore suggested fitting the solar cells to theXI-V for testing and decided to go ahead with testing as part of a joint research project.

We replaced the solar cells on one side of the satellite with newly developed CIGS solar cells, made from a compound semiconductor consisting of copper, indium, gallium and selenium. We also affixed expensive gallium arsenide solar cells to three of the other sides, in return for conducting tests. We affixed the same general-purpose silicon solar cells that we had used previously to the remaining two sides, resulting in an unusual mixture of three different types of solar cell. We also improved the performance of the camera in an effort to calculate the satellite’s position and movements via continuous shooting, as an additional mission.

As the XI-V was to be released fromSSETI Express, we conducted a number of tests in the Netherlands at the European Space Research and Technology Centre (ESTEC), whereExpress was being developed. To start with, two students went over in 2004 with a flight model for theXI-V, to make sure that the release mechanism could definitely hold and release the satellite.

We brought the satellite back for the time being and then in summer of 2005, another student took the completed theXI-V over again to be fitted. As engineers at ESTEC were to handle everything from that point through to the launch, all we could do was to carefully check the operation manual from start to finish and hope that they didn’t make any mistakes.

A secret gamble

The launch was originally scheduled for July, but was put back by three months due to issues with the rocket. The revised date was October 27. Just a few days prior to the launch however, something unexpected happened. ESTEC contacted us and reported that a problem had arisen with the lid of the release mechanism for theXI-V. The lid was held shut with fishing line, which was supposed to be cut by a nichrome wire when a current was passed through it from a battery. As the capacity of the battery had been reduced however, it seemed to be taking around a minute for the wire to heat up and cut through, rather than several seconds as intended. We had to decide whether to push ahead regardless or put our plans on hold.

If we wanted to be 100% certain of success, it would have made sense to put our plans on hold. If we let this chance slip by however, we didn’t know when we might get another. We weren’t keen on having to repeat the whole process from filling out paperwork through to adjusting the interface either. After a heated debate with graduate student and Project Manager Funase Ryu (now at JAXA), we concluded that, even if the fishing line weren’t cut to begin with, it would eventually break due to the high levels of cosmic radiation in space. We agreed “let’s take a gamble on that” and gave the green light to go ahead. We also decided not to tell the other students about our conversation, and just kept our fingers crossed right through to the day itself.

While Funase and I were keeping our concerns under wraps, the rocket carryingSSETI Express and theXI-V took off on October 27, 2005, from the Plesetsk Cosmodrome just like the previous launch. I was out of the lab on the day because I was testing another satellite, but I got back to the university in time for the first pass over the skies above Tokyo and sat there with the students waiting for the radio waves from theXI-V.

It started to pass over at 9:40am, but we couldn’t pick up the beacon. Perhaps the fishing line hadn’t broken after all, preventing theXI-V from being released. Both Funase and I had that same unspeakable thought. If that was indeed the case, then all we could do was wait for the fishing line to break naturally. Just as I was making up my mind however, the second pass started, 80 minutes later. This time we managed to receive the beacon signal with no problem. TheXI-V had been successfully released and was flying through space.

It is unclear why we could hear the beacon during the second pass but not the first. The most likely explanation seems to be that the fishing line had almost been cut at the time of the first pass but not completely. Then by the time of the second pass, it had broken naturally, causing the lid of the release mechanism to open and eject the satellite.

TheXI-V may have got off to a rocky start, but we managed to identify its orbit after that and it successfully began its mission as planned. Orbit testing of the new CIGS solar cells is going particularly well and continues to produce results that are making Kawakita very happy at JAXA, including the world’s first consecutive data for long-term generation. We ran a comparison on the three different types of solar cell in fall of 2010, five years on from the launch. Whereas the generation efficiency of the other two cells has deteriorated compared to initial levels, the CIGS cells have retained their performance relatively well, proving that they are highly resistant to radiation. This shows that an inexpensive, quickly developed nanosatellite is the ideal means to quickly test a newly developed technology in space.

On to practical missions

When we completedXI-IV in 2002, we were eager to put the knowledge and experience that we had built up developing a CubeSat to good use. As we wanted to develop a more practical nanosatellite with earth observation capabilities, we decided to launch a project to develop a new satellite. CalledPRISM, the aim of the project was to see what level of surface resolution we could achieve from a university-developed nanosatellite. Our goal was 30m. Whereas medium and large-scale satellites normally use metallic reflective optical systems, we developed a new refractive optical system that used softer materials to make it lighter and more compact.

One of the most important technologies we used was a lens tube that expands like a telescope. This combined extremely flexible “super elastic” materials with a rigid frame. Whereas there would usually be a support in the middle, this causes problems because it can obstruct the view. The system that we developed on the other hand ensures a clear view. We have recently obtained a patent.

In addition to a high performance optical system, we also developed a range of other new technologies to achieve our primary mission, including a magnetic torquer that changes position magnetically, a solar censor-based precision attitude control system, a transmitter with a new high capacity modulation system, and a computer equipped with multiple CPUs. Our satellite was chosen by JAXA ahead of other applicants to piggyback on itsH-IIA launch vehicle, and passed Japanese safety reviews, which are regarded as the toughest in the world. We were finally able to launch on board a Japanese rocket.

In January 2009,PRISM was launched into orbit. Once the satellite had been released and had moved into the correct position, we obtained its position information, we stopped it rotating and then extended the lens tube. We tuned the optical system and started to take images of clouds and Earth from around 10 weeks onwards. We then analyzed the resulting images and confirmed that we had achieved a surface resolution of 30m.

The next challenge that we set ourselves was to develop our first ever fully-fledged scientific satellite, in the form of our astrometric satelliteNano-JASMINE. In order to accurately measure even faint stars, up to eighth magnitude, it is necessary to restrict any variation in the position of the satellite to within 0.7 arcseconds (approximately 1/5000 of a degree) for a period of eight seconds. This requires an exceptionally advanced level of technology. We had already completed a flight model and were working on the final launch adjustments. It may have started out as an educational project, but we are now trying to establish CubeSat technology as a world-leading science.

Making space accessible to everyone

In 2010, the Japanese government designated the “development of a new paradigm in space development and utilization by nanosatellite,” including the Japanese concept of “reasonably reliable system engineering,” as proposed by myself and my colleagues, as an advanced research and development support program. This translates into investment totaling ¥4.1 billion over the next five years. As the name suggests, “reasonably reliable system engineering” is a form of engineering concerned with ensuring reasonable levels of reliability, rather than overreaching and having to deal with shortages as a result.

The biggest advantage of micro/nanosatellites is that they can be developed quickly and inexpensively. The rigid designs and in-depth testing required for medium and large-scale satellites are unnecessary and would actually detract from that. What we want to do is to clearly spell out what is required to ensure a reasonable level of reliability, whilst making the most of those benefits in terms of speed and cost. With that in mind, we would like to create a pool of micro/nanosatellites that would be accessible to anyone, so as to harness the elemental micro/nanosatellite technologies developed by the likes of universities to date. We would also like to develop and share inexpensive testing methods and development processes for use on micro/nanosatellites.

We are planning to develop and put into operation five micro/nanosatellites as part of this program in the future. First up, we intend to make an Earth observation satellite with a high resolution of 5m. We will then publish our observation data so that they can be freely used by companies, universities and other organizations.

Our ultimate aim is to get universities and smaller companies throughout Japan involved, thereby creating a micro/nanosatellite community. Instead of individual researchers having to make satellites for their own purposes, we want to make micro/nanosatellites available to schools, smaller companies and local communities so that they can find new ways of using them. We need to significantly expand the range of uses for satellites, and come up with a steady stream of applications specific to micro/nanosatellites. Only then will micro/nanosatellites genuinely become public property and take root as a key component of space exploration.

Translated from “Uchuu kaihatsu: Uchuu ni te ga todoita,” Nikkei Science,September 2011, pp. 4–47. (Courtesy of Nikkei Science Inc.)