Modernizing Manufacturing: How to Build the Satellite of the Future
Satellite manufacturing today is a lengthy, meticulous process; its high tech nature, and the cost in time and money make advances slow compared to sectors like the mobile industry. But an explosion in demand for connectivity and other space services is driving the need for ever-more capable satellites. It is at this crossroads, that bold new ideas are being forged.
The satellite industry is a very patient industry. A typical geostationary spacecraft takes two to three years to build, and lasts well over a decade — sometimes two. By comparison, automotive companies can, with all parts on hand, build a car in less than a day. Fancier cars, like a Rolls Royce, may take a few months, but the point is clear: other tech industries move much, much faster.
Despite the lengthy production times, satellites are far from obsolete, and an ever-increasing number are parked above the Earth’s atmosphere every year. Still, from cars to smartphones, there is a desire to make satellite manufacturing more akin to other high tech fields of today.
So how do satellites stay competitive? According to Martin Halliwell, CTO of SES, terrestrial communications providers have always outpaced the satellite industry, with the main reason being how risky it is to try new ideas on multi-million dollar assets that, once powered up, are too far away to fix if something goes wrong.
“Those expectations are not the highest because we are generally a very conservative crowd due to the fact that after we have launched a spacecraft, we have no opportunity to modify it or change things, so we tend to change things very slowly. It seems to be more an evolutionary process than a revolutionary process,” he says.
Halliwell says the satellite industry maintains its value by providing services from a vantage point that terrestrial telecommunications companies simply cannot. Ubiquitous coverage of large geographies, connectivity in the air and at sea, certain forms of secure government communications, and even the ability to cost-effectively distribute Ultra-HD content are all areas where satellite has the high ground.
But this does not mean that the satellite industry can, or is, resting on its laurels. Stagnation is death, and to sit idle would inevitably mean losing market share to something else. Halliwell says there have been gradual improvements in different facets of satellite production, such as more efficient solar arrays, better sensors, and the use of photonics. What he describes as a true game changer, however, is the development of highly flexible payloads.
Building it Better
“What we would like to see is the ability to be able to receive the signal on the spacecraft, break it up immediately by hertz so every single hertz of usable bandwidth can be routed where we want it to be, and to who we want it to be used by. That degree of flexibility is not there today — it’s nowhere near there today. So there is a lot of work to be done,” he says.
Satellite manufacturers are privy to these demands, says Mark Spiwak, president of Boeing Satellite Systems International. Greater flexibility with bandwidth is one of the top new requirements the company is seeing in Request for Proposals (RFPs) by satellite operators.
“For example, let’s say you have the World Cup or the Olympics coming up. There will be a surge in demand for bandwidth in that particular area as people want to tune in and watch, have access to Internet, etc. Satellite customers want the flexibility to shift bandwidth to this area, to move it to capture the needs of their customers,” he says.
Spiwak says Boeing has been, and continues to invest in such technology today. Intelsat 29e launched Jan. 27 as the first commercial satellite with a Boeing-built digital payload. Boeing originally developed the technology for the U.S. Department of Defense’s (DoD) Wideband Global Satcom (WGS) satellites. Spiwak said the company is already researching and developing its seventh generation digital payload for the future.
Digital payloads are not the only new technology in demand either. Spiwak lists higher throughput capabilities, lower cost per gigabits per second (Gbps), and increasing schedule pressure to bring new products and spacecraft to market faster as other rising expectations found in RFPs. To expedite the build-process and meet these demands, component suppliers are stepping their game up as well.
“Commercial telecom, at least a decade ago, wasn’t maybe feeling a lot of change pressure, but it seems like in the last few years in particular commercial telecom — and now maybe even to some extent our institutional customers — are feeling a lot of schedule and price pressure,” says Michael Pavloff, CTO of Ruag Space.
Ruag is finding that more of its customers are interested in trying new methods of producing satellites. Much of this momentum, says Pavloff, stems from entrepreneurial players that want less expensive satellite equipment, and want it now.
“This whole ‘NewSpace’ world, I think it has really driven some radical change, which is starting to fold back into the rest of industry … customers are demanding electronics that can be a factor of 10 or more less expensive on much more aggressive delivery schedules versus what we see from our legacy markets,” he explains.
Pavloff says Ruag Space is tapping into techniques developed outside of the satellite industry to meet these demands. In addition to serving the satellite telecommunications industry and institutional customers like the European Space Agency (ESA), the company also performs lots of industrial high-volume work for non-space industries. Pavloff says Ruag has been manufacturing electronic components for other industries at much higher rates, and now as demand grows for higher production, shorter lead times, and the use of Commercial Off the Shelf (COTS) components, those skills are transferring to the space world.
Spiwak mentions a similar strategy at Boeing, where there is now an increasing focus on bringing technologies and techniques from other industries such as airplane manufacturing into satellite production systems. Both Spiwak and Pavloff point to increased robotic assembly and automation as ways their companies are optimizing the way satellites are constructed. These methods remove touch-labor, shortening the time and effort needed to complete rote tasks.
“To manufacture a sandwich panel, which historically was done with hand drilling and hand-potting of inserts into the panels, we have invested in a robotic capability called the automatic potting machine to be able to do this insert manufacturing for the future,” says Pavloff. “There have been some technology developments associated with this, not just the robotic aspects, but mass reduction technology with new inserts and adhesives as well.”
Automation is also seeing greater use in the testing procedures used to vet spacecraft prior to launch. “We have designed our digital payload with the ability to test itself, eliminating the need for additional tests or additional test equipment,” adds Spiwak.
Any discussion on spacecraft manufacturing today would be remiss not to note the impact of small satellites. Halliwell says the difference between small satellites and large geosynchronous satellites is very wide to the point where he sees little crossover. A major factor is how customized geosynchronous satellites are compared to their smaller, non-geosynchronous brethren. Though very different in mission and design, Spiwak says SmallSats are pushing Boeing to look at new technologies, production systems and engineering methodology. One very tangible impact is on the next generation of satellite manufacturers, since CubeSats and other small projects are increasingly a part of academic learning.
“Most of the new engineers Boeing has hired in the last five years have already worked on real satellites in college — they speak the language, know the algorithms and hit the ground running, which is adding tremendous energy and speed to our developments,” he says.
Because of the increasing capability of small satellites, new entrepreneurs are seeking out ways to use them for telecommunications — shooting for greater success than their 1990’s counterparts. Brian Holz, CEO of OneWeb Satellites, says his company — a recently formed joint venture between OneWeb and Airbus Defence and Space — aims to crank out two to three satellites everyday once fully operational. He stresses, however, that these satellites should not be confused with CubeSats or other similar microsatellites. The 150 kilogram spacecraft OneWeb Satellites plans to build will have more than 6 Gbps of throughput, accurate pointing capability, and a power to weight ratio similar to a traditional telecom satellite.
Holz says much of the trick to mass-producing satellites so quickly will reside in supply chain management. At the beginning, OneWeb Satellites has a development program that will take a couple of years, he says, and will start with two qualification satellites, followed by a 10-satellite pilot program. Once warmed up, Holz says the company will ramp to its desired production rate over five or six months, ultimately reaching the multiple-satellites-per-day clip around two to three months into OneWeb’s launch campaign.
“On the production side, it’s really key how we manage our suppliers and supply chain management, our integration and assembly processes, overall industrialization approach, our overall test approach, etc., to do production at timelines that aren’t typically seen in industry,” he explains. “Essentially we will have delivered long lead items on the shelf and then our production timeline starts.”
Pavloff says collocation is a growing trend for Ruag Space that is reshaping the traditional component supplier-satellite manufacturer relationship. Ruag Space is bidding on roughly 25 percent of the OneWeb Satellites spacecraft.
“Closer partnerships with our customers have become increasingly important. This has taken on multiple aspects, one of which is our customers, whether we are talking about an ESA project or OneWeb, have started to find it advantageous for all parties to have a collaborative engineering process up front. Co-engineering has become a critical part of what we do,” he says, adding that sometimes this involves risk and profit-sharing partnerships that make Ruag Space more of an economic partner.
Holz says the bulk of the effort is going to be scrutinizing the industrial process and automating that from a space perspective. The goal is for each satellite to have a five-year design life with a 90 percent probability of success. Since OneWeb has ordered 900 satellites, the company will have “safety in numbers” to build in redundancy and hedge against the occasional satellite failure. Holz says many satellites will last longer than five years too.
“We are taking the time on the front end to make sure we design for manufacturing, so it’s looking at how the system interfaces together and is assembled quickly, pushing tests down to lower levels of assembly and doing less at the end. We have a very robust test program at our box level and then module level. Our system has a good bit of modularity for its size, and we do a lot of our yield management and burn-in testing at a modular level before it’s assembled at a system level,” he says.
Building Satellites in the Next Decade
Satellite manufacturing may be a rigid process today, but visions for the future are much more dynamic. Halliwell says in the 2020 to 2025 timeframe, SES is planning satellites that are much more nimble and adaptable. Not only does this mean flexible payloads, but a new way to augment satellites even after they are launched.
“We are going to launch satellites with a capability to be able to add subsequent payloads once they are in orbit and on station. This is something we discuss in general terms. We are working with a particular manufacturer at the moment to develop this to the next stage. The idea is that each of the satellites we launch in that timeframe will have on the back of the spacecraft a docking mechanism. It will have an electrical and mechanical interface to allow us connect at a later date, another payload,” he says.
A rising number of companies are studying adaptable hardware for spacecraft to enable such capabilities. NovaWurks, for example, is designing its Hyper-Integrated Satlet (HISat) with the ability to conform around payloads and perform other self-assembling tasks in orbit.
One company working to possibly upend the modern conception of satellite manufacturing is Made In Space. Founded in 2010, the company is pioneering the use of 3-D printing for on-orbit manufacturing.
“While other people were looking at solutions like reusable rockets or alternatives to chemical propulsion, we asked a very different question: what if you didn’t have to launch at all? What if you could build things in space? That was the genesis of the idea,” says Jason Dunn, co-founder and CTO of Made In Space.
In 2014, Made In Space shipped a 3-D printer to the International Space Station (ISS) and proved that additive manufacturing is possible in space. The next step is starting commercial services with a more robust device: the Additive Manufacturing Facility (AMF), which launched to the ISS in March on an Orbital ATK resupply mission. Dunn says the AMF will enable Made In Space to build satellites in orbit with partner NanoRacks. The service, called Stash & Deploy, constructs satellites using the AMF and “stashed” components stored on the station.
According to Dunn, 3-D printing satellites in orbit means they can be built without “over-engineering” to survive launch, or with size restrictions to fit inside a payload fairing. He says 3-D printing and robotic assembly can one day build spacecraft that are much more complex than those of today.
Beyond Stash & Deploy, Made In Space is working with NASA on an even bigger project called Archinaut — short for architecture astronaut — that can build whole structures in space.
“What Archinaut enables is an ability to do in-space robotic manufacturing and assembly that could build entire space systems,” he explains. “Some variations of Archinaut build sub-systems for spacecraft, and other variations have the ability to modify or augment spacecraft on orbit.”
Additive manufacturing draws mixed opinions across the satellite industry, with some seeing great potential, while others view it as an incremental benefit. Pavloff says Ruag Space used 3-D printing to build an S-band antenna bracket with 50 percent less mass and with improved structural property in roughly half the time and cost. He expects Ruag Space to use more 3-D printing in specific situations, such as where low mass is preferred. Boeing uses additive manufacturing for some parts of its 702 satellite product line, and Spiwak says the company plans to include more 3-D printing in the future. For OneWeb Satellites, Holz says 3-D printing is interesting, but requires too large an upfront capital investment to be of interest at this stage.
Dunn expects 3-D printing in space to first draw the allure of academia, but then pick up within the overall space industry. As far as what can be built, Dunn says the possibilities span the entire gamut of spacecraft in orbit today and beyond.
“We looked at the whole range from CubeSats to mega-satellites — ones that are much bigger than we are deploying today. I’m probably more interested in the mega-sized satellites, because that is where we really get into a new design philosophy. This means building things on orbit that overcome the limits of portable systems today. For us that means things like antennas that you can manufacture on orbit and get to a larger aperture size than we are accustomed to,” he says.
Over the next two years, Dunn anticipates flight missions coming to fruition, and then more projects much like Archinaut transitioning out of R&D and becoming a more public part of the company. Projects like these, he says, could herald a new era in space. VS