The Space Reviewin association with SpaceNews

AsiaSat 3 lunar flyby
Artist’s rendering of AsiaSat 3 / HGS-1 during a lunar flyby, 1998. (credit: Hughes Global Services)

Beyond GEO, commercially: 15 years… and counting

Bookmark and Share

Fifteen years ago today the AsiaSat 3 spacecraft became the first spacecraft in the history of space travel to conduct commercial operations in the vicinity of the Moon, executing the first of two successful lunar flybys designed to take it from an initial useless orbit to a final usable one.

By mid-1998, nearly five months after being stranded in an odd orbit around the Earth and five weeks after its first lunar swingby, the spacecraft was finally parked in a slightly inclined geosynchronous Earth orbit, where it soon completed initial checkout for providing telecommunications services.

Before this mission, only government space agencies of the US, USSR, and Japan had attempted robotic lunar missions. Approximately 100 were launched and over half were total failures.1

Sadly, fifteen years later, this mission still stands as the only commercial mission ever conducted well beyond the Geostationary Earth Orbit (GEO) arc.

The AsiaSat 3 salvage mission happened nearly thirty years after the first serious proposal had been floated to conduct a commercial mission at the Moon (Project Harvest Moon), and its success capped the 20th century with a milestone of particular importance to commercial space advocates expecting significant expansion of commercial “deep space” mission activity in the early 21st century.

Sadly, fifteen years later, this mission still stands as the only commercial mission ever conducted well beyond that special place where the bulk of all current commercial space activity happens—and ends: the Geostationary Earth Orbit (GEO) arc. This ring around the Earth high above the Equator is a virtual parking lot for hundreds of commercial telecommunications satellites and wellspring of approximately two-thirds of the nearly $300-billion global space economy.

As far as commercial space goes, however, the GEO arc has served as a metaphor for the edge of the known world, with the Moon, near-Earth asteroids and Mars beckoning far across the abyss, untouched by commercial interests.

This brief recap of the first commercial venture beyond the GEO arc highlights how such achievements can result not just from methodical advanced planning, detailed engineering, and preparation as one might expect and hope for, but also from a wild, unpredictable mix of failure, serendipity, personal motivation, passion, calculated risk-taking, scrappiness, compromise, connections, insider knowledge, ignorance, egos, and corporate politics—as well as old-fashioned good luck.

Note: The conversations recollected below are accurate summaries of what was said, and not necessarily word-for-word transcripts.

To the rescue

The AsiaSat 3 mission started as a failure. In late December 1997, the final stage of its Russian Proton launcher malfunctioned early into the second of two planned burns at apogee, leaving the spacecraft in a highly inclined, highly elliptical Earth orbit—a Geostationary Transfer Orbit (GTO) in mission-design parlance. Mission controllers from the spacecraft’s manufacturer, Hughes Space and Communications in El Segundo, California, readily took control of the errant satellite. Obscure headlines and brief, matter-of-fact press releases reported the failure, but it generally escaped much notice, even from me, due to the busy holiday season.

About six months into a new job at Microcosm, Inc. (Torrance, California), I was managing its division focused on space-mission engineering and analysis. I had left JPL after an exciting 11-year stint to pursue a passion for pushing commercial space activity beyond Earth orbit and into the solar system.

For about four months leading up to the 1997 holiday season, I was actively working with Microcosm founder and CEO Jim Wertz to get a potential technical and business collaboration in place between Microcosm and another small space-mission engineering firm, Innovative Orbit Design (Boston). IOD was founded and led by Ed Belbruno, a creative mathematician and trajectory theorist with whom I had collaborated in the early 1990s on low-cost lunar mission concepts when we were both working at JPL.

In 1986, Belbruno conceived of a novel and propellant-efficient orbit-transfer technique for getting spacecraft from the vicinity of the Earth to the Moon. His “Fuzzy Boundary” or “Weak Stability Boundary” (WSB) technique2 was attempted some five years later in 1991 by the Japanese Hiten spacecraft. Hiten had dropped off a small lunar orbiter (Hagoromo) during a lunar swingby in 1990, but the small subsatellite’s transmitter had failed. Hiten, Japan’s first lunar probe, carried on with its mission—multiple lunar flybys and the first-ever aerobraking maneuver, at Earth—and then slid into lunar orbit after successful application of the WSB transfer approach. This fortuitous opportunity salvaged the lunar orbiting phase of the Hiten-Hagoromo mission and validated Belbruno’s theoretical work.3 Subsequent to this dramatic demonstration, Belbruno sought and was granted several patents relating to these novel orbit-transfer techniques.

typical WSB trajectory
A typical WSB trajectory leading to a lunar orbit capture, much like that used on the Hiten mission.

Belbruno had visited Microcosm in late October that year and discussions about the proposed collaboration were progressing well. Wertz’ stated focus for Microcosm was “finding alternative technologies to reduce space-mission cost,” so a Microcosm-IOD partnership made sense.

During the first two weeks in January after the holidays, letters and conference calls addressing the details of the proposed collaboration were exchanged between Wertz and Belbruno’s representatives on the east coast (two attorneys), while I was preparing detailed task plans for the staff at Microcosm.

Sometime after the New Year one of Belbruno’s attorneys had been tipped off about the AsiaSat 3 failure, and soon Belbruno was aware of these events and pondering the situation. Belbruno called me on January 12th, wondering if I had heard about the failure. He noted that if the spacecraft was still healthy, his WSB technique might be able to save this mission as well—by sending AsiaSat 3 to the Moon first. I replied that I had seen the headlines but paid little attention, and knew next to nothing else about the spacecraft or its condition—but also knew how to find out.

My space career started at Hughes in the late 1970s, and during my five years there supporting five space missions I had worked with some of the best spacecraft system and mission engineers on staff, and had kept in touch with a few over the years.

I ran some back-of-the-envelope calculations assessing whether there was enough propellant onboard to follow Belbruno’s guidance, send the vehicle to the Moon, crank the craft’s orbit using one or more lunar flybys, fling it back to GEO, and salvage the mission.

First on my call list was Loren Slafer, who always seemed to know what was going on across the company. By 1998 he was Hughes’ Chief Technologist for spacecraft attitude determination and control systems, well aware and connected, but I hadn’t talked with him for about four years. I reached him on January 16th and, after a few pleasantries, asked if he could tell me anything about the state of AsiaSat 3.

“Nothing in detail, but the mission team is working on various scenarios for what might be done,” Slafer said. “Most of the detailed information is embargoed due to an ongoing insurance investigation.”

He gave me the number of Hughes mission engineer Chris Cutroneo, with whom I had worked on all five missions back in the mid-1980s but hadn’t talked with for over a dozen years. After catching up on news a bit, we got down to business.

“Chris, Loren tells me you’re working on mission options for AsiaSat 3. What can you tell me about the state of the spacecraft? I may have a suggestion for you.”

“What kind of suggestion?” Chris asked.

“Well... it involves applying an unconventional orbit-transfer technique to move the satellite to a better place,“ I answered, avoiding mention of the Moon altogether.

“Sorry Rex, I can’t tell you anything specific,” he said. “The insurers are all over this—and us—trying to figure out what went wrong, and the Hughes lawyers are advising us not to say anything to anyone.”

“Anything? You can’t say anything?” I asked. “All I need to know are basic numbers like masses, propellant loads, its current orbit, and whether the spacecraft is healthy and under control.”

“Well,” Chris said, “if you happen to go online and happen to come across a certain summary of AsiaSat 3, especially what’s on the Hughes website for the HS-601HP spacecraft model and in the recent AsiaSat press releases, all of that is probably pretty accurate info.” (Wink, wink, nudge, nudge.) “And I can say that we’re controlling the spacecraft, it’s healthy, and all we have done so far is raise its perigee by about 150 kilometers. Beyond this, I can’t say what were up to, but we’re actively assessing our options.”

“Thanks Chris. I think this is all I need,” I said.

Three weeks had passed since AsiaSat 3 was left stranded in Earth orbit. The insurers had already written it off as a total loss and had committed to make good on the $200-million policy. They were also paying the Hughes mission analysts to figure out what could be done with the vehicle that might recoup some of this sizable loss—someday.

Crunching numbers

Using some basic rule-of-thumb WSB maneuver numbers from Belbruno, spacecraft specs found online for AsiaSat 3, and the current orbit as reported in the press (plus factoring in the 150-kilometer perigee tweak), I ran some back-of-the-envelope calculations assessing whether there was enough propellant onboard to follow Belbruno’s guidance, send the vehicle to the Moon, crank the craft’s orbit using one or more lunar flybys, fling it back to GEO, and salvage the mission. To cover my bases, I was able to independently confirm some of the AsiaSat 3 specs by e-mailing some questions directly to AsiaSat in Hong Kong, and Cutroneo allowed me to query a member of his Hughes team who was familiar with odd trajectories, Cesar Ocampo.

Ocampo had completed some graduate-level intern work at JPL a few years earlier and was familiar with Belbruno and his novel trajectories, so I threw out a hint to Ocampo that the WSB approach might prove useful if applied to AsiaSat 3’s predicament. Ocampo was intrigued—but I still didn’t mention the Moon.

Convinced from my simple analysis that things looked promising, I asked another Microcosm mission analyst who reported to me, Curtis Potterveld, to take a crack at the same calculations and double-check my results.

“I’ve been running all of the proposed orbits here since the failure, and nobody has uttered the word ‘Moon’.”

Curtis came to the same conclusions I did, which were so amazing that we asked another Microcosm analyst, John Collins, to check both of us. John made his own refinements to the calculations, modifying our numbers a bit and leading to slightly better results.

On January 19th, a week after Belbruno’s call, I summarized our findings in an internal memo to Wertz:

“It appears that given the current conditions of the HS-601HP spacecraft (esp. the huge biprop propellant load) and its current ‘useless’ orbit (GTO), it could easily be inserted into lunar orbit!”

Our numbers also confirmed Belbruno’s intuition about using the Moon to get AsiaSat 3 back to its intended parking spot: we estimated that after a three-month trip it could be stopped at GEO with enough onboard propellant remaining to support at least several years of orbit-slot stationkeeping operations.

I called Cutroneo mid-morning to convey our findings. “Chris, we’ve completed our initial assessment of AsiaSat 3’s situation,” I said. “Are you sitting down?” I summarized the WSB approach and our numbers—and mentioned going to the Moon.

“Wow, Rex,” he said. “You need to call Cesar, right now!”

I replayed our findings to Ocampo, and he immediately comprehended our suggestion.

“I’ve been running all of the proposed orbits here since the failure, and nobody has uttered the word ‘Moon’.”

Cesar and Belbruno started talking shop.

Reality sets in

During the next six weeks, a flurry of activity on all fronts dramatically transformed the notional mission concept conceived by Belbruno, validated by the team at Microcosm and lobbed by me into the option studies at Hughes into an executable salvage mission plan.4

Also during this period, most of the collegial relationships that had been formed to work the opportunity gradually unraveled and degraded, leaving most of the involved parties unaware—or at best partly aware—of what the others were doing. Attitudes soured and frustrations grew.

Ocampo’s enthusiasm for Belbruno’s WSB suggestion quickly transferred to Cutroneo, who then informed his AsiaSat contacts of this compelling development while seeking internal approval at Hughes to work the idea into their ongoing mission studies. Pending this approval, Cutroneo signaled on January 20th that he wanted to start having face-to-face mission-design meetings ASAP.

Wertz continued to work with Belbruno’s attorneys to put finishing touches on the proposed Microcosm-IOD collaboration, while Slafer and I tried to stay on top of daily developments.

Slafer called me on January 22nd and reported that a quick-look assessment by the Hughes team of sending AsiaSat 3 to the Moon resulted in no show-stoppers and that the mission team was “very excited” about the opportunity and wants to do it, adding that “I think the insurers would want to try it anyway, regardless of what Hughes thinks.”

Slafer estimated that the effort was going to cost about $1 million to implement (mostly to cover special engineering and analysis tasks), and suggested avenues the Microcosm-IOD team might consider for approaching Hughes with a proposal to lend technical support to the salvage effort.

Late in January, Belbruno’s team contacted the AsiaSat insurers directly, informing them of the salvage opportunity and seeking their support for the idea. (We learned about five weeks later that Hughes management was quite irate about this at the time, but they eventually calmed down.)

Salvatore had concluded that the WSB approach was untenable for the AsiaSat 3 situation, and was instead working out the details for trying an Apollo-style “free-return” trajectory.

Slafer called me on January 28th with an update—and the first big reality check. He reported that following additional technical assessment, the Hughes engineers had determined that there actually was one show-stopper to applying Belbruno’s WSB salvage approach, but only one: due to radio link performance margins, mission controllers would not be able to communicate with AsiaSat 3 much farther out than the Moon’s distance (approximately 400 thousand kilometers), while Belbruno’s trajectory required the craft to use a lunar swingby to slingshot out as far as nearly four times the Earth-Moon distance (approximately 1.5 million kilometers), where subtle gravitational influences from the Sun would alter the parameters of the craft’s orbit in desirable ways.

Whereas during January most developments on the salvage effort were relatively open and trackable, events during February and March played out more in a blur of multiple, uncoordinated communications channels, mixed signals, and stealthy developments. All of us involved had our own view of the Big Picture, much like the various blind men feeling the elephant. None of us had a clear view of what was actually going on.

I was having occasional conversations and phone calls with Wertz and Belbruno’s team about the Microcosm-IOD partnership, while Wertz separately was having more frequent detailed follow-up calls with Belbruno’s attorneys and even the AsiaSat 3 insurers—but terms of the deal were not converging very well. In fact, Wertz asked me to stand down with my efforts until these deal points were sorted out, but since events were moving along at such a rapid pace I (regretfully) didn’t follow his suggestion. I was relying on phone tag-ups with Slafer to stay informed about developments at Hughes, which we had during work hours as well as after hours. Belbruno was engaged in a series of technical calls with Ocampo, and Ocampo was also briefing colleagues internally at Hughes about how IOD’s WSB technique might be applicable to many future Hughes comsat missions. Belbruno’s attorneys were separately trying to convince Hughes management to strike some sort of strategic partnership and collaborative R&D effort with IOD. His attorneys were even calling me after hours to discuss strategy and sort out plans. The insurers were calling IOD with encouraging updates, claiming in mid-March that “Hughes is going to get this salvage opportunity.” And the Hughes lawyers continued to remind those involved with the salvage to keep their lips zipped.

By late March, Slafer had concluded that based on what he knew that not only was a Hughes-IOD collaboration off the table, but so was using any sort of WSB-based technique for the salvage. Belbruno confirmed this on March 27th, reporting to me that Ocampo had just debriefed him on an internal Hughes mission team status meeting held the week before.

“They’re going to use a free-return trajectory,” he said, sighing.

In the dark

Unknown to any of us outside of Hughes at the time and not discovered until late April, since early February (at the latest) Hughes’ Chief Technologist for Orbital Mechanics, Jerry Salvatore, was working an alternative salvage plan involving the Moon. (Nearly two years later at a talk given at JPL,5 Salvatore said that the first suggestion to consider using the Moon in salvage scenarios was passed to him by another Hughes colleague in one of the company parking lots.)

Salvatore played a major role is all four prior Hughes salvage missions—Anik C, SBS, Leasat 3, and Intelsat VI, all missions salvaged using the Space Shuttle—and was one of the best at his craft and very clever.

Salvatore had concluded that the WSB approach was untenable for the AsiaSat 3 situation, and was instead working out the details for trying an Apollo-style “free-return” trajectory, which still involved sending the craft to the Moon for a flyby, but cranked the orbit there using the gravity-assist effect, rather than flinging the spacecraft into deep space so the Sun could work on it.

Reference 4 below by Ocampo provides an excellent technical summary of the salvage mission trajectory design and other technical details. Suffice it to say that given the time and constraints within which the Hughes team had to devise their salvage plan, the final trajectory chosen and key events required to pull it off, while not optimum, were “good enough.”

In retrospect, through most of April those of us outside Hughes had little idea what was really going on inside Hughes. It was even worse than February and March.

Nearly everything relating to the salvage effort that Wertz, Belbruno, and I were working on at that time seemed in limbo. The Microcosm-IOD partnership process was stalling out, as were the ongoing Hughes-IOD collaboration discussions. Ocampo was still in a dialog with Belbruno, and on April 23rd reported that things at Hughes with regard to the salvage were “still looking fairly good” and that a decision on what to do with AsiaSat 3 was coming down the pike. “There is nothing negative at all,” he told Belbruno.

Then things got ugly.


On April 27th, Belbruno called me, extremely agitated (with a few choice words omitted here).

“Rex! Hughes is going to the Moon with AsiaSat 3!”

He had contacts in Europe, some tied to the amateur astronomer community, who during the previous few days were hearing rumors from Paris indicating outward movement of AsiaSat 3, toward the Moon.

No mention was made whatsoever at the press conference or in the press releases about Belbruno’s role in all of this, IOD, my role or Microcosm, nor were any of us notified in advance about the event.

Via the Internet and email, these rumors were rapidly confirmed and spread, reportedly forcing Hughes to hold a press conference to announce their plans earlier than they had intended. This was held on April 29th, accompanied by press releases from Hughes as well as Analytical Graphics, Inc., whose orbital mechanics software was used in part to help devise the salvage trajectory.

At the press conference, another Hughes subsidiary, Hughes Global Services, announced that they had assumed title of the errant spacecraft and that it had been renamed HGS-1.

The first maneuver committing HGS-1 to the salvage plan had actually occurred on April 10th, and the Hughes team was gradually raising the craft’s apogee higher and higher, with a plan to commit to the first of two planned lunar flybys with a trans-lunar injection burn scheduled for May 7th, followed by a lunar swingby on May 13th.

trajectory side view
Side view of key trajectory events for salvaging AsiaSat 3 / HGS-1. (credit: Hughes Global Services)

No mention was made whatsoever at the press conference or in the press releases about Belbruno’s role in all of this, IOD, my role or Microcosm, nor were any of us notified in advance about the event. This made Belbruno apoplectic and was personally dismaying to me. Nonetheless, I was delighted that there was finally a commercial spacecraft heading to the Moon.

Most media channels picked up the story, and it was front-page news the next day, April 30th. Coincidentally, this was the same day that I was on the program at the biennial Low-Cost Planetary Missions Conference (held at Caltech in Pasadena) to present a paper6 co-authored with a Microcosm colleague specifically addressing the history of and prospects for commercial deep-space missions. The timing couldn’t have been more perfect, and I milked it for all it was worth.


During the next ten days or so, Belbruno and I tried our best to learn more about what was happening by contacting Slafer and Ocampo. We also initiated our own PR campaign by compiling our version of how this saga started and e-mailing it to our own media contacts, space colleagues and chat rooms. (Remember chat rooms?) On May 5th Wertz submitted a letter setting the facts straight to the editor of the local South Bay paper, the Daily Breeze, and I faxed a copy to Slafer for good measure. (It was published verbatim in the Breeze on May 13th, the day of HGS-1’s first lunar flyby.)

I finally got hold of Slafer on May 8th, and he was upbeat. To his credit, he and Cutroneo had marched over to senior management in Hughes’ PR department the day before to summarize the facts as they viewed them, and Slafer waved his copy of Wertz’ letter to the Breeze and confirmed that everything in this letter was accurate. He also enthusiastically offered to get approval for a suggestion I made during the call: to allow Belbruno and me access to the Hughes Mission Control Center in El Segundo during the upcoming historic lunar flyby.

Thanks to the efforts of our front-line champions inside Hughes, our many supporters in the space community, and the connectivity of the Internet, the story about how the core, enabling idea of using the Moon to salvage AsiaSat 3 entered into the option trade space at Hughes was not buried by the personal or corporate motivations that apparently wanted to squelch the facts.

On the same day I called the MCC contact Slafer suggested, and she saw no reason why the request wouldn’t be approved. But on May 11th she called me back and reported that the MCC would be closed to “all but Hughes employees” during the encounter. (I confirmed months later that several non-Hughes people were in the MCC during the encounter, which by this time didn’t surprise me.)

Ocampo got back to me on May 12th, lamenting that his management didn’t want him talking with anyone on the outside about how the salvage is going “until it’s all over.” Fair enough. But then he added, “I have my own opinions about who did what when, but my management doesn’t want me to talk about those.”7

The May 13th lunar flyby went off without a hitch, as did the second flyby on June 6th. Ten days later HGS-1 was parked in the desired slightly inclined (8 degrees) geosynchronous orbit. Normal orbital perturbations would gradually decrease the inclination toward zero.

trajectory top view
Top view of key trajectory events for salvaging AsiaSat 3 / HGS-1. (credit: Hughes Global Services)

Planned and executed in a record four months, the first commercial mission beyond GEO was a complete success. Commands were sent by ground controllers to deploy the twin solar arrays (one didn’t) and large telecom antennas (all did), and HGS-1 was ready for service.

After a brief marketing effort, HGS sold the satellite to PanAmSat the next year. It was then moved to a different orbital slot in GEO and successfully operated for three more years. Nearly out of propellant, it was deactivated in July 2002 and moved to a graveyard orbit.8

Thanks to the efforts of our front-line champions inside Hughes, our many supporters in the space community, and the connectivity of the Internet, the story about how the core, enabling idea of using the Moon to salvage AsiaSat 3 entered into the option trade space at Hughes was not buried by the personal or corporate motivations that apparently wanted to squelch the facts. The story got out and stuck.

For the annual Aviation Week & Space Technology magazine Laureate Awards, arguably the closest thing aerospace has to the film industry’s Academy Awards, Salvatore, Ocampo, Belbruno, and Ridenoure were nominated for the 1998 Best Achievement in Space award. Though from a handful of nominations the ultimate Laureate award went to NASA Administrator Dan Goldin, just being nominated speaks for itself.

What next?

With the success of the AsiaSat 3 salvage on the books, similar salvage opportunities have been taken seriously since. Using the Moon to help salvage a stranded spacecraft—all GEO comsats—was discussed and worked to some level of technical detail for Orion 3 in 2000, and again in 2006 for ArabSat 4A and 2008 for AMC-14. But for various reasons salvages were not attempted.

There are currently at least a dozen active commercial or private efforts intending—on a planned, not contingency basis—to develop and launch a spacecraft or space system beyond GEO to destinations afar. Some target the Moon, some near Earth asteroids, others Mars, and others heliocentric orbits between Venus and the Earth or various libration points in cislunar space. Many are frequently making headlines and creating buzz.

Unless there is a very mature program operating in stealth mode right now, the earliest launch date for any of these systems seems to be 2014 or 2015. Until then, the tally for commercial space activity beyond GEO stands at one—15 years ago and counting.


1 For a complete robotic lunar mission list, see

2 For an introduction to low-energy orbit transfers such as the WSB approach, see, which also includes links to several additional useful references.

3 For more background on the Hiten / Hagoromo mission, see

4 2005 December, C. Ocampo: “Trajectory Analysis of the Lunar Flyby Rescue ofAsiaSat-3/HGS-1”; Published in the Annals of the New York Academy of Sciences, Volume 1065 New Trends in Astrodynamics and Applications, Pages ix–x, 1–470. We are all indebted to Cesar Ocampo for thoroughly documenting much of what happened technically and programmatically inside Hughes during the period covered by this article. I referred to this reference and my own voluminous notes extensively when compiling this chronicle.

5 1999 November 30, J. Salvatore and M. Skidmore: “The HGS-1 Salvage Mission”; Caltech Management Association lecture, held at JPL.

6 1998 April 30, R. Ridenoure and K. Polk: “Private, Commercial and Student-Oriented Low-cost Deep-Space Missions: A Global Survey of Activity”; Paper for 3rd IAA International Conference on Low-Cost Planetary Missions held at Caltech from April 27-May 1.

7 See p. 234 in Reference 4 for Ocampo’s frank summary of what was going on behind the scenes. There is much more that could be said and written about the backroom shenanigans relating to this saga, including litigation, but that would be like wanting to see how sausage is made, wouldn’t it?

8 For the PAS-22 summary, see