Smallsats need small propulsion. Boston startup Accion has a few big ideas.


This article originally appeared in the July 3, 2017 issue of SpaceNews magazine.

U.S. military and commercial satellite operators are eager to hand off some of the jobs large spacecraft perform to small satellites and even cubesats. Before that can happen, though, small satellites need their own propulsion systems because most of the widely used chemical and electric propulsion technologies don’t fit well on shoebox-size satellites and they are difficult to scale down.

Natalya Bailey is well aware of this problem. While working toward her PhD at the Massachusetts Institute of Technology, Bailey and fellow graduate student Louis Perna developed and prototyped tiny microsatellite thrusters that MIT later delivered to the Aerospace Corp. in Los Angeles for integration in two cubesats, called AeroCubes.

By the time the AeroCubes were launched, Bailey and Perna had founded Accion Systems, a Boston-based company focused on electric propulsion systems for small satellites.

Since 2014, Accion has raised $10.5 million in venture capital and $6.5 million from unnamed U.S. Defense Department agencies.

Bailey spoke recently with SpaceNews correspondent Debra Werner.

Natalya Bailey, co-founder of Accion Systems. Credit: Accion Systems
Natalya Bailey, co-founder of Accion Systems. Credit: Accion Systems

What made you think the time was right to found Accion?

I tried to start another microsatellite propulsion company around 2008. Around then, and it remains true for Accion, there was starting to be a lot of pull and demand from the market.

When I was at MIT working on this technology with my cofounder, the Department of Defense was very interested in it and was funding some of the early research. Some of the big primes were either trying to buy flight systems from MIT or trying to license the intellectual property. A pull from the market was our main sign.

It seems like that pull is allowing you to go from laboratory test to flight quickly.

That’s right. The promise of small satellites and cubesats hinges on being able to scale down a few key systems like propulsion and power. Not everything follows Moore’s law. So these gaps are getting a lot of attention right now and pushing us to move faster.

When do you expect to test your technology in flight?

The technology itself has flown a couple times when we were still at MIT. In terms of proving out the technology on orbit, that’s been done with MIT and the Aerospace Corporation and my cofounder and myself when we were still grad students. We are delivering our first flight hardware this fall for a first quarter 2018 launch.

What spacecraft will that fly on?

I would love to say but can’t yet.

What’s unique about your technology?

We are building what is essentially an ion engine, but there are two main differences. We use a liquid propellant and completely new manufacturing techniques. The liquid propellant allows us to produce thrust and create ion engines on a very small scale. Our smallest complete ion engine is about the size of a pack of cards. More traditionally they’ve been the size of a Prius. That’s all due to using liquid instead of a pressurized neutral gas like xenon or argon.

As for the manufacturing techniques, aerospace was previously characterized by very bespoke, one-off processes. Given the trends in the industry, we thought it would be wise to look to other applications and manufacturing techniques across industries and pull from some that are mature and amenable to high-volume throughput to drive down our costs.

A dime-sized thruster chip developed by Accion Systems. Credit: Accion Systems
A dime-sized thruster chip developed by Accion Systems. Credit: Accion Systems

Like what?

We’ve adapted some of the same processes used to make computer chips in silicon to make our engines. For other components, we rely on proven processes like conventional machining. We basically tried to find the most mature, standard, mass-producible processes that we could and used those.

Are those processes designed to keep costs low enough for cubesats?

Yes. As we are starting out, the costs for some of our first runs are not as low as they will be in the future so it is easier to partner with folks with slightly larger satellites. We are targeting everything from three kilograms up to about 200 kilograms with our products. There is definitely always going to be a race to the bottom in terms of prices for cubesat propulsion.

Are you on a cubesat for your first flight?


A lot of new propulsion technologies are coming into the market. 

Yes. When we started out, people would ask about the competitive landscape. There was no one except Busek. Now we hear about a new competitor every week. It’s a great motivator and we are focused and moving quickly. It’s definitely a popular area.

Do you think that once you fly there will be more demand for your technology?

Yes and no. The people building and operating satellites have risk profiles all over the place today. For a certain batch of them, propulsion is critical to their business model. They are ready to adopt something that seems like it will work once you launch it.

On the other end of the spectrum, there are people who want to see things fly for five or 10 years before they will adopt them. We are on the former part of that spectrum initially. I don’t think flying will change the market, but it will let us address a wider swath of customers.

What was challenging about miniaturizing the ion engine?

With traditional designs, the way you generate thrust is by injecting a neutral gas, say xenon, into a chamber. Then you inject a beam of high energy electrons into the same chamber and you have to guarantee that those electrons will collide with the neutral atoms and create ions. You accelerate those ions out the back of the chamber to produce thrust.

But as you start trying to decrease the size of the chamber, all of those particles you inject in will leave much more quickly before they have a chance to collide and create an ion. You end up just emitting neutral gas and electrons out the back and not producing any thrust.

The way to solve that is to inject more and more high energy electrons to guarantee that you will create ions from that neutral gas. The problem is a materials one. You inject more and more high-energy electrons and you start melting the walls of any material known to man that’s manufacturable and that could be launched into space. There is a point where you couldn’t shrink it down anymore because it would heat up too much and would melt the material.

Rather than trying to solve that problem, we looked at a completely different way of producing ions. We take a liquid propellant that is just positive and negative ions and apply a strong electric field to the liquid and pluck individual ions out of the liquid and accelerate them out the back.

Does the flight heritage on AeroCubes make your technology unique?

You have to differentiate between chemical or cold gas and electric propulsion.

It is quite easy to launch a can of air on a cubesat, open it and get little puffs of thrust. You may need more than one hand to count the number of times that has been done. As far as efficient electric propulsion where you can actually perform operational missions, there are very few propulsion technologies that have flown.