Putting eyes in the sky on the cheap

Award-winning expat Chris Boshuizen, co-founder of earth observation company Planet Labs, is about to tackle his next challenge. Brent Balinski spoke to Boshuizen about the excitement around space start-ups, the agile aerospace approach to satellite manufacturing, and whether or not Australia needs a space agency.

“The world’s changing… and to think you can get some money from your parents and beat the Soviet Union!” said Chris Boshuizen, co-founder of space business success story Planet Labs, with a chuckle.

“To be honest, for $75,000 for the launch and about $5,000 in bits, including your phone, you can put a satellite in space. And every family that’s reasonably okay off can mortgage their house and have their own space programme.”

The world is, of course, a very different place compared to what it was in 1957, when the Sputnik was launched. No longer something that requires the resources of a USSR or US, space can be conquered by bootstrapped enterprises with nifty ideas. And investors are excited.

Research from market intelligence firm CB Insights released in June found $US 1.167 billion invested in space start-ups for the year to that date (largely anchored by Elon Musk’s SpaceX). This was more than the previous three whole years combined.

“Ten years ago space start-ups didn’t exist – now we’re not even sure if we’ve thought of all the ideas yet for them,” Boshuizen told Manufacturers’ Monthly this week.

Boshuizen – who was last year’s Advance Global Australian of the year (and champion in the same awards’ Advanced Manufacturing category) – had his interest in space piqued at age three or four via Lego and Carl Sagan’s Cosmos. It was the one night a week where a 7 pm bedtime didn’t apply.

“I remember that quite profoundly as an important thing if I was allowed to stay up, so it was probably pretty formative,” he said.

Following a doctorate at University of Sydney in Physics and a stint at NASA’s Ames Research Center, Boshuizen and two other engineers from ARC started Planet Labs in 2010.

At Ames, Boshuizen and co-founder Will Marshall were goaded by engineering director Pete Klupar to put phones in space, after being repeatedly told that Klupar’s BlackBerry was more capable than satellites currently operating. This led to the PhoneSat project, which NASA still runs.

The Doves are always watching

Boshuizen and Marshall left the space agency to pursue the idea of disrupting the space industry through low-cost, off-the-shelf electronics and a software-like approach to building hardware.

As of October, the enterprise has attracted $US 183 million in venture capital backing and has a timetable to have 150 shoebox-sized satellites in orbit 475 km from the earth in the third quarter of next year.

The satellites, which the company calls Doves, have a 3U CubeSat form factor, a two- to three-year lifespan, weigh about 4 kilograms, and use a lot of components found in smartphones.

They are “always on”, imaging the earth, and remain static as the earth spins, acting – as the company likes to say – like a line scanner for the planet. (See diagram.) Other imaging satellites are generally task-based.

The company aims to image the entire earth once a day, every day, with the data of enormous value in tracking things like crop yields (an agribusiness firm is among its customers), deforestation, and the impact of natural disasters.

Doves are assembled in-house, with rapid iterations (enabled by similarly rapid improvements in things such as sensors) and a “release early, release often” approach often seen among Silicon Valley companies. The satellites are in their 13th generation.

The agile aerospace approach works due to things such as the cheapness of the craft and the distributed, low-level risk of failure (two launches have exploded, though the lost Doves were replaced easily enough). It would not be an option for designing, for examples, old-school satellites (such as a LandSat costing hundreds of millions of dollars) or a passenger jet.

“So if you start at the other end and you iterate a lot then you can learn a lot really quickly,” he explained.

“Sort of a requirements-driven process puts a lot of the learning at the end, because you don’t know what you’ve built until you’ve built it. And you’ve got to build everything based on requirements and guessing.

Walking through the history of manufacturing

Boshuizen describes the approach in terms of scaling as somewhere between old-style satellites – in batches of one every few years -and mass-produced electronic goods.

“A gen one Motorola prototype might be 1,000 a gen two prototype might be 5,000, and the final release prototype might be 10,000,” he explained.

“We still want some  of the benefits of scale, but it doesn’t make sense to be making 10,000 in a run, right?

“I like to joke that a good analogy is like Ford Motor Company circa 1924 or something, where they’ve got the production line and they’re benefiting from parallelisation and batch processing and people doing simple jobs, but it’s not fully robotic, automated, Toyota circa 1980. So we’re sort of walking through the history of manufacturing from the 1900s.”

The company had its first Doves launched from the International Space Station in January 2014. They hitch a ride into space wherever there’s room, on rockets including Ukrainian Dneprs, Orbital Science’s Antares and SpaceX’s Dragons. Over 100 Doves have been launched, with around half still orbiting.

Boshuizen left his post as CTO in October, though is retained as a mentor and adviser.

He is currently examining where the next opportunity might lie in the space industry. Boshuizen said there were ample prospects in small satellites, in many areas that require “lots of eyes”, though added the trend of internet satellite ventures was something that excited him less.

“There’s a lot of applications; earth imaging is one, and where is the next Planet Labs? And so I’m currently doing a research project to answer that question,” he said.

“I’ll probably be announcing something in December so it’s a bit hush-hush at the moment.”

Do we need a space agency?

The PhoneSat project has spawned around 20 start-ups, reckons Boshuizen. It continues a long list of technology emerging from NASA and creating economic benefit, a function outlined in the 1958 Space Act.

According to a count in 2012 by the organisation’s publication Spinoff, some 1,800 innovations have transferred into the US economy. These are as varied as memory foam, aviation anti-icing systems, and safety grooving for roads.

Asked whether Australia would benefit from a space agency, for benefits including developing and commercialising new ideas, Boshuizen said a 23-year-old him enthusiastically argued the case, but things are different now.

“No matter which country you’re in – if you’re in the US it’s ‘let’s get NASA more money’, if you’re in Sweden it’s ‘let’s start a Swedish space agency’… that’s the first place people go,” he offered.

“Which is kind of like saying ‘I wish someone else would solve my problem.’”

Plus, nowadays you don’t need the might of a well-funded government agency to get to space.

“Australia needs people to get off their butts and do something,” he said.

“And so what Australia really needs is 20 innovative space companies and then Australia will be a leader in space. And that’s purely in the hands of the people who want to start that.”

Images:

Planet Labs

sbs.com.au

NASA