For a few brief minutes around 1:40am Tokyo time on April 26, the world was allowed to believe that a tiny start-up might have pulled off an engineering miracle 384,000km from Earth.
Ispace appeared to have made Japan only the fourth country to land a spacecraft on the Moon and itself the first private sector enterprise to do so.
A livestream of the landing cut out as the Hakuto-R lander was on its final descent and it was not until around 20 minutes later that Takeshi Hakamada, the company’s chief executive and public face of ispace, acknowledged that the lander had probably not touched down.
He nevertheless declared the mission a “huge achievement”. The reality lay strewn in pieces across the dusty surface of the Moon. The screens could not show it, but in its attempt at a giant leap for Japan, ispace had stumbled.
Viewers were warned that ground controllers were likely to lose contact with Hakuto-R at an unknown height above the lunar surface. But they were not aware that some seasoned space engineers closely involved in the project were astonished that it had even got that close.
More than five decades after man first set foot on its surface, missions to the Moon are back in fashion, with South Korea, Canada, Mexico and Israel all planning to send rovers to explore it. India last month became the fourth country to land a probe on the Moon.
But for all its prowess in technology, Japan has yet to crack one of mankind’s greatest technological challenges. Ispace offered the chance to change that.
Exploring space is technically challenging and expensive, and lunar landings are especially difficult. But an investigation by the Financial Times suggests that the Hakuto-R failure was more than just an unfortunate crash. Its demise followed months of financial, technological and corporate turmoil at one of the most ambitious national projects ever entrusted to Japan’s private sector.
This account is based on court and internal company documents alongside interviews with dozens of former and current ispace employees — who all spoke on condition of anonymity for fear of retribution — plus company executives, bankers, technology advisers, investors, business partners and space industry executives.
Many described a corporate environment within ispace where failure was not allowed and working conditions were “toxic”. In some cases, technology concerns were allegedly pushed aside amid pressure from shareholders, lenders and business partners to get the first mission off the ground.
The turnover of engineers at ispace was so high that entire teams sometimes left at once. Ispace declined to comment on specific allegations, but said in a statement that its values included diversity and that employee concerns were taken seriously.
Ispace sold itself to investors as a dynamic private-sector operator in the mould of Elon Musk’s SpaceX or Jeff Bezos’s Blue Origin by talking of business opportunities on the Moon. Four months after its first lunar mission ended in failure its shares — owned by thousands of individual investors — still trade at more than six times their listing price.
But funding for commercial space ventures is in decline. Some ispace employees worry that if the start-up cannot prove it can one day make money by carrying payloads to the Moon’s surface, then investors will take flight.
“Maybe it’s not so hard to us to succeed with landing on the Moon if we build on our experience,” says one Japanese employee. “The bigger challenge is how we sustain our business from here. That’s the biggest question I have.”
Asking for the moon
Hakamada, who has a masters degree in aerospace engineering from Georgia Institute of Technology, started ispace in 2010 while taking part in a $30mn lunar landing competition set up by Google.
That contest ultimately ended without a winner but Hakamada continued his pursuit of space travel, raising $90.2mn for ispace in late 2017. The fundraising, the highest ever in a series A round in Japan, brought in heavyweights such as the state-backed Innovation Network Corporation of Japan (INCJ) and the country’s most influential venture capitalist, Tohru Akaura, alongside advertising giant Dentsu, Japan Airlines, telecoms company KDDI and more recently Katsunori Sago, SoftBank’s former chief strategy officer.
Its main lender was Sumitomo Mitsui Financial Group, whose brokerage arm SMBC Nikko later underwrote ispace’s IPO, while MS&AD Insurance Group, another investor, provided “the world’s first lunar mission insurance contract”.
Hakamada started with building a rover, intended to move around the moon’s surface. But he shifted to developing a lander — a kind of lunar delivery vehicle to ferry payloads from an orbiter to the Moon. The 44-year-old enthusiastically sold his vision, dubbed “Moon Valley 2040”, to investors, telling them that by 2040 the moon would be inhabited by 1,000 people, with 10,000 visitors each year. “Until the launch, ispace was a marketing company,” says one former executive.
In a statement, ispace said: “As a commercial enterprise to maintain a sustainable business model, a robust marketing element and financing capability are essential.”
“We did think it was a pretty wild investment case when we were doing our due diligence,” admits Tatsuhiko Nishimura, managing director at INCJ. “But we credited the management for being able to access the right people and the key individuals in the industry.”
Ispace recruited engineers from around the world, who were attracted to the so-called “new space” sector spearheaded by SpaceX and including around 10,000 other commercial space companies.
The company’s pitch was to drive down costs and bring innovation to space technologies that would take decades to develop under state-backed programmes. Instead of developing technology from scratch, ispace presented itself as an “integrator” of established knowhow. It has a licensing deal to use guidance, navigation and control (GNC) software developed by Draper, the US company behind the guidance systems for the first manned landings on the moon in 1969.
For its lander, it procured the propulsion system from France’s ArianeGroup. “Ispace had drawn up the scenario that its first mission would succeed by combining its system integration technology with the technologies drawn from Draper and ArianeGroup,” says Kazuya Yoshida, a prominent space robotics expert from Tohoku University who has advised ispace on early phase technology.
But the gap between ispace’s vision and reality widened as it grew from an organisation with just a handful of staff to more than 200 employees, of whom 60 per cent come from outside Japan.
The company initially struggled to find engineers with actual experience in the space industry and its first mission was delayed by four years. From 2020 until last summer, the development of the missions was overseen by a chief technology officer who previously worked at Sony and Panasonic and had no direct experience in space.
Ispace began to attract more high-quality engineers, but the management soon faced a different challenge: a stream of “power harassment” or bullying claims. Abusive behaviour by some team leaders, instances of discrimination against non-Japanese engineers and incompetence among certain managers were all alleged.
In one case, alleged bullying by a single Japanese team leader led to at least 10 people leaving before the individual was demoted last year, according to three people familiar with the matter.
“It wasn’t just that the working environment was toxic but the management tolerated that,” says a former engineer who left recently. “Apart from the toxicity, loads of people had the feeling that they didn’t belong and there was no sense of a team that was ‘all in this together’.”
In June, Kyle Acierno, ispace’s former head of its US business, filed a lawsuit in a Colorado federal court, alleging that the Japanese company discriminated against him by dismissing him last year because of his race and national origin. According to the court filing, Acierno alleged that Hakamada justified the termination by telling him that the Japanese were “afraid of Americans”. Ispace is due to respond by September 8. Acierno declined to comment.
In a statement, ispace declined to comment on the Acierno litigation, the high turnover and other specific allegations of bullying, abusive behaviour and discrimination. But it added that its values include diversity and that has helped the company’s growth. “Any allegation of harassment or bullying is taken seriously and handled officially through the appropriate HR process.” The company has since hired Ron Garan, a former astronaut, to head its US subsidiary.
Discontent was widespread, with some Japanese employees complaining they were paid less than non-Japanese colleagues.
Another major source of employee frustration was the frequency with which management made technical decisions that went against the advice of engineers — either to save costs or to satisfy demands from investors.
“We were essentially behaving like a bank rather than a space company,” says one former team leader. “Decisions were often made to cut on time, money and resources. Some of us were thinking that if anything, they were going to throw a rock because they just needed to show that they’re launching something.”
Ispace responded: “To achieve our vision of an Earth-Moon economy, we work to involve traditionally non-space companies and the general public . . . This programme has served ispace and partner organisations well and provided us the opportunity to undertake our challenging lunar missions”.
Shares have lift off
In March 2023, as its most valuable asset was hurtling towards the Moon aboard a SpaceX rocket, ispace surprised both its employees and the industry by pushing ahead with a listing on the Tokyo Stock Exchange.
While Bezos and Musk prefer to keep their space businesses in private hands, Hakamada tells the Financial Times that “the understanding of a greater number of people” was needed to build a new space industry. “Rather than raising private money, we wanted to become a company that can be assessed by capital markets.”
Jumpei Nozaki, ispace’s chief financial officer, adds that there was no funding shortage or banking pressure behind its decision to pursue an IPO, but analysts point out that Japan lacks the pool of risk capital that would, in the US or European context, support a company through late-stage funding.
“A lot of Japanese start-ups are certainly doing IPOs far too early in their evolution as companies, but there is not a lot of choice for a company like ispace,” says one adviser to the space industry.
Banks were hardly lining up to underwrite the offering. According to four people with knowledge of the talks, Morgan Stanley was among several that considered taking part in the IPO, but pulled out after assessing the risks. Nomura, Japan’s biggest brokerage, and another US bank also decided against participation. All three banks and ispace declined to comment.
The prospectus for its $50mn IPO stood out for a list of risk factors that spanned nearly 30 pages. Sole bookrunner SMBC Nikko asked investors to sign a document confirming that they were aware of the risks of investing in ispace, according to a person with knowledge of the situation. The company declined to comment.
In November, ahead of both the IPO and the original launch date, ispace announced 10 “mission milestones”. They were devised after employees asked management for concrete success criteria, but the slick way they were presented raised eyebrows among suppliers and engineers. Only the final two involved actually landing on the Moon, prompting some to think the company already had the possibility of a failure in mind.
“Each milestone is very significant and each represents an important success,” Nozaki says. “We may have failed with milestone 9 and 10 but the success of milestone 1 through 8 is a big achievement.”
Despite the risks and bankers’ caution, the offer was popular with individual investors. Ispace was priced at ¥254 per share and began trading on April 12. Because the IPO price was 78 per cent below the valuation achieved at its final private funding round, the shares took off, peaking at over ¥2,000 shortly before Hakuto-R was scheduled to touch down on the Moon.
The postmortem
An internal analysis by ispace blamed a software glitch for the lander’s demise.
Ahead of the mission, the company had changed the proposed landing site from relatively flat terrain to the 88km-wide Atlas crater. Draper’s software was deceived by the lander’s passage over the rim of a crater, computing that it was near the surface when in fact it was roughly 5km above. Descending at a low speed for a soft landing, the lander ran out of fuel and crashed.
Hakamada says the main issue was with Draper’s software but acknowledged the fault ultimately lay with ispace since it had not adequately adjusted the software’s algorithms to reflect the changed landing site.
For many employees, the biggest question was why landing algorithms treated the Moon’s surface as flat when its craters can be seen from earth. According to one engineer, the algorithms were initially simplified to assume the terrain was flat but that decision was not revisited after ispace chose a new landing site.
A person close to the company denies that the Moon’s surface was assumed to be flat and said the glitch occurred because the simulation range for the lunar terrain was reduced.
One veteran engineer also says the company “did all kinds of testing for satellites . . . but there was nothing at all related to landing, just verification that the sensors and landing algorithms were working. That was completely done by Draper and the integration was validated in a simulation.”
In a statement, Julia MacDonough, general manager of space systems at Draper, said the company and ispace’s team for GNC software development provided progress reports on schedule, and as requested. “Draper performed the verification of our algorithms and software as directed by ispace,” she added.
Ispace engineers privately say they would have preferred to do testing in a mock environment. “It didn’t feel right that everything could be done by simulation,” one says.
“No matter how much we talked to the management, it was clear that they had neither the intention nor the capacity to do the testing,” the engineer adds. “There was no room for debate about whether that was the right stance or not.”
Hakamada says it was “mostly meaningless” to do landing tests on earth, given its much stronger gravity, but adds that some would be carried out ahead of a second mission. Ispace said: “Our teamwork with Draper was and is strong”.
The launch went ahead anyway. A current engineer with knowledge of the internal discussions says there was “an atmosphere where we could not say ‘let’s delay the mission for another six months because the simulation is not enough’.”
Several engineers closely involved with the mission say they expressed concerns to Hakamada months before the failed landing. Some put the mission’s chances of success at “between 25 per cent and 50 per cent”.
“We honestly felt that landing would be extremely difficult and we thought the mission would be a success if we even came close to the Moon,” says one ispace engineer who monitored the mission from inside the control room.
The debris from the lander is now scattered on the lunar surface. Hakamada is undeterred, however, saying ispace is using “massive money” and “making all of our plans in order to succeed”. Another mission to the Moon is planned for 2024, though it is possible that an attempt by Japan’s national space agency, which blasted off on September 7, will get there first.
“But at the same time, we cannot avoid all of the risks,” he adds. “So we need to make our business plans with that certain degree of risk in mind.”
Read the full article here