≡ Menu

Sprites: A Chip-Sized Spacecraft Solution

In mid-June, NASA announced the award of two contracts with Deep Space Industries in conjunction with the agency’s plans to work with private industry in the exploration and harvesting of asteroids. One of these contracts caught my eye immediately. It involves small payloads that can ride along to supplement asteroid missions, and it’s in the hands of NASA’s former Chief Technologist, Mason Peck, a Cornell University aerospace engineer. Peck’s work at Cornell’s Space Systems Design Studio has led to the development of Sprites, fully functional spacecraft each weighing less than a penny. You can think of a Sprite as a spacecraft on a chip without any constraints from onboard fuel.


You can see where this fits in with the current theme of building smaller spacecraft and sending them in swarms to investigate a particular target. You may have already run into KickSat, a citizen science project involving hundreds of proof-of-concept spacecraft in low Earth orbit for assessment of their performance and re-entry characteristics. KickSat grew out of a KickStarter campaign from 2011. The diminutive spacecraft are 32x32x4mm in size, each weighing less than 7.5 grams, designed to be released from the larger KickSat, a CubeSat modified and enhanced for Sprite deployment, on command from the ground.

Image: Aerospace engineer Mason Peck, whose Sprite concept shrinks spacecraft to the size of micro-chips. Credit: NASA/Bill Ingalls.

KickSat was launched on April 18th of this year, the plan being to release more than 100 Sprites, which would have become the smallest satellites ever to orbit the Earth. Unfortunately, the KickSat satellite reentered the atmosphere without Sprite deployment, leading to talk of building KickSat-2. The latest KickSat-2 update from Zachary Manchester, a member of Mason Peck’s lab at Cornell, is here. But as the new satellite takes shape, let’s talk about those Sprites. For while the KickSat experiments could provide broad spatial coverage of near-Earth phenomena, there is nothing to prevent the use of sprites to create sensor nets for deep space.

Modes of Propulsion

In Exploring Space with Chip-sized Satellites, an article in IEEE Spectrum in 2011, Peck explained that radiation pressure from the Sun offers one way for Sprites to move around the Solar System. They’re too small for onboard propellant, but the ratio of surface area to volume ensures that they can be driven just like a tiny sail. Peck explains the idea in relation to a much larger sail, the Japanese IKAROS:

If a Sprite could be made thin enough, then its entire body could act as a solar sail. We calculate that at a thickness of about 20 micrometers—which is feasible with existing fabrication techniques—a 7.5-mg Sprite would have the right ratio of surface area to volume to accelerate at about 0.06 mm/s2, maybe 10 times as fast as IKAROS. That should be enough for some interplanetary missions. If Sprites could be printed on even thinner material, they could accelerate to speeds that might even take them out of the solar system and on toward distant stars.


Image: Size of the Sprite satellite. Credit: Space Systems Design Studio.

Earlier this week we looked at Jordin Kare’s work on SailBeam, a concept involving vast numbers of tiny ‘micro-sails’. The Sprite has an affinity with Kare’s thinking, but unlike Kare, who was going to drive his microsails with a multi-billion watt orbiting laser, Peck is also exploring how charged Sprites might interact with the magnetic fields that surround planets. The Lorentz force bends the trajectory of a charged particle moving through a magnetic field. Can we put a charge on a Sprite?

In his lab work at Cornell, Peck and colleagues have tested ways of exposing Sprites to xenon plasma, mimicking conditions in Earth’s ionosphere. The Sprite can use a power supply to put a potential between two wires extending from the chip, letting plasma interactions charge the device. The charge is maintained as long as the Sprite continues to power its wires, so we can turn it on and off. If we can manipulate the charge aboard a Sprite at will, then imagine exposing a stream of charged Sprites to Jupiter’s magnetic field, 20,000 times the strength of Earth’s.

Jupiter as particle accelerator? The idea seems made to order, particularly since we’ve been examining particle accelerators of a vastly different order of magnitude — remember the 105 kilometer accelerators we talked about in relation to Cliff Singer’s pellet propulsion concepts. The nice thing about Jupiter is that we don’t have to build it. Here we have a way to accelerate one Sprite or 10,000 of them to speeds of thousands of kilometers per second, at which point the chips could shed their charge and be flung off on an interstellar journey.

Peck adds that getting the Sprites up to speed might itself take decades, and the journey to the nearest star would still be a matter of several centuries. But 300 years to Alpha Centauri beats any solar-sail-plus-Sundiver-maneuver mission I’ve ever seen, and unlike the admittedly faster beamed lightsail missions (some of Forward’s missions get down to decades), the Sprites take advantage of a form of propulsion that doesn’t require vast infrastructure in space.

Near-Term Issues

We’re talking, of course, about future generation Sprites, tiny spacecraft that have been built to surmount the problems Peck’s team is now trying to solve. Take the issue of damage along the way, which we had to think about both with Cliff Singer’s pellets and Gerald Nordley’s self-steering ‘snowflake’ craft. Better build many and be prepared for some losses. Lightweight Sprites have no radiation shielding, leaving the electronics vulnerable, and micrometeorites within the Solar System pose their own threat. The way to overcome such problems in the near-term is to send Sprites in large numbers, assuming a degree of loss during the mission.


Image: Artist’s conception of a cloud of Sprite satellites over the Earth. Credit: Space Systems Design Studio.

For missions deep into the Solar System and beyond it, though, we have to solve these problems. But I love the idea of using sunlight or the Lorentz force to accelerate these tiny payloads, which also have a natural synergy with CubeSats. Remember that The Planetary Society’s LightSail-1 is testing sail deployment from CubeSats, potentially creating a way to deliver a CubeSat laden with Sprites to other planets in the Solar System. Before we think of scaling to interstellar, why not think in terms of legions of Sprites sending back data from the surface of Mars, or placed into orbits that could provide deeply detailed maps of the solar wind and flare activity?

As we do this, we can be learning how best to deploy future Sprites, and how to fabricate everything from spectrometers to load sensors and basic cameras on a chip. Peck notes in the IEEE article that almost everything a spacecraft has to do can be managed with semiconductors, from solar cells for power, capacitors for energy storage and the various requirements of memory and processing. Take these ideas down to much smaller scales and the idea of swarm probes exploring the outer planets begins to resonate, with obvious implications for the kind of payloads we will one day want to send to Alpha Centauri.


Comments on this entry are closed.

  • ljk July 17, 2014, 14:20

    Links to my related CD articles on this piece.

    Deep Space Propulsion via Magnetic Fields


    ChipSat: To the Stars via Magnetic Fields


    Putting the Pieces Together in Space


    Reconfigurable Structures in Space: Q & A


  • ljk July 17, 2014, 14:39

    “The nice thing about Jupiter is that we don’t have to build it.”

    Now there is a quote. :^)

    Jupiter as a starship launcher. Hmmmm. And it has some ready-made bases of operation to boot (the moons).

  • Rob Wisniewski July 17, 2014, 16:09

    Jupiter as a particle accelerator is a ‘mind blown’ moment for me. Great article.

    Any concerns about the ability for such a small craft to communicate effectively? Surface area to volume ratios and charge-based propulsion make for great proportion-related efficiency in maneuvering, but unfortunately the energy needed to send information over distance doesn’t change with the size of the craft. What’s our lower limit on being able to store and use enough energy to communicate?

  • Andrew Palfreyman July 17, 2014, 16:18

    For me, the mandatory precursor to any attempt at interstellar travel has to be Maccone’s gravscope. Look before you leap. Getting to 550+ AU fairly quickly and in multiple locations seems an ideal mission for swarms of sprites. There’s a ton of unresolved engineering issues, but as a big picture it’s not a bad start.

  • Wojciech J July 17, 2014, 16:25

    Such small observational devices that we can develop already lead me to inevitable question…what would a civilization developed millions of years ahead of us be capable of of if it wanted to remain undetected. Even with our technology we could have observational sensor nets in the Oort Cloud that would remain undetected for Earth telescopes and astronomers.

    On another topic connected to this-I wonder if we could combine swarm-like behavior(already being tested in drones) probes and hypertelescope concepts in such probes? A swarm of millions of micro-telescopes in focal point of Solar System’s gravitational lens? Could it be possible or are they too small?

  • Wojciech J July 17, 2014, 16:29

    Rob Wisniewski July 17, 2014 at 16:09

    ” What’s our lower limit on being able to store and use enough energy to communicate?”
    I wonder if we could develop Queen Bee concept with such probes. Basically for certain number of probes you would have a much larger probe using beamed energy to fuel others and allow them to communicate? Not every probe could have the same configuration, there could be specialists(communication, energy supply, transport, repair etc..)

  • norman wells July 17, 2014, 17:45

    Is there not enough debris orbiting Earth already? Why do we not develop the means of clearing what is already in orbit before we add to it!

  • Kamal Ali July 17, 2014, 18:13

    How about sterilized sprites for sampling the Enceladus or Europa vents?
    They could be dispersed gradually so we could get a good sample from many points within the venting regions.
    They would probably only be powerful enough to transmit to an orbiter, but we’d get a sensor net of data.

  • Eniac July 17, 2014, 20:54

    Solar sailing and Lorentz acceleration are exciting and apparently quite feasible modes of propulsion for sprites, and the most exciting thing is that there are no consumables, i.e sprites can have lifetimes of years or even decades during which they can travel continuously.

    As Rob Wisniewski observes, though, the fly in the ointment is communication. The power requirements of interplanetary communication are such that sprites will be incommunicative, for all practical purposes. One way out I can see is if you send out beams of them, so that the distances between them stay small and communications can be relayed all the way back to the origin. This will require vast numbers, though, and good flight coordination in order to not let large swarms of them become isolated. It will also mean much larger latency than that provided by direct radio.

    Another way out could be laser beaming. With highly collimated semiconductor or fiber micro-lasers, you would have a much more energy efficient way to communicate. I don’t think the technology is there, but it is conceivably not far off.

    Of course, this problem is magnified million-fold if we are thinking about interstellar applications.There is no way we could stay in touch with interstellar sprites, they would have to be completely self-sufficient.

  • Eniac July 17, 2014, 21:00

    There is one other option regarding sprite communications: A large swarm of sprites could act cooperatively to form a microwave phased array capable of sending and receiving with extremely high gain. So, if you have enough of them in one place, you may not need a trail of breadcrumbs, you may be able to communicate by microwave pencil-beam instead. Even, conceivably, at much larger scale over interstellar distances.

  • Paul Gilster July 17, 2014, 21:23

    Norman Wells writes:

    Is there not enough debris orbiting Earth already? Why do we not develop the means of clearing what is already in orbit before we add to it!

    The KickSat Sprites would all have burned up in the atmosphere as their mission ended — that was part of the plan. On a related note, NanoSail-D was originally developed as a way of de-orbiting larger satellites, so work is going into this issue.

  • Michael July 18, 2014, 0:37

    One day we may be able to clear Jupiter’s radiation belts and inject our own particles, now that would make a powerful accelerator. But back to the sprites, the radiation would easily penetrate their structures and fry them!

  • swage July 18, 2014, 5:34

    I like it. Miniaturization is very clearly a key factor for Interstellar endeavors because of involved energy need. The more practical experience we get with unconventional approaches, the more we will enable our capabilities for future projects. Certainly the experience will influence and change the “big” projects on the horizon into something more practical. Absolutely invaluable.

  • Mark July 18, 2014, 11:52

    This brings up one possible avenue of search for solving the Fermi paradox. If we’re able to construct space probes a few cm in size, stands to reason that advanced civilizations should be able to construct them at the molecular level.

    Instead of training our telescopes on the skies, perhaps we should be training our microscopes on the dust.

  • ljk July 18, 2014, 13:50

    Mark said on July 18, 2014 at 11:52:

    “This brings up one possible avenue of search for solving the Fermi paradox. If we’re able to construct space probes a few cm in size, stands to reason that advanced civilizations should be able to construct them at the molecular level.

    “Instead of training our telescopes on the skies, perhaps we should be training our microscopes on the dust.”

    And thus the reason for this Centauri Dreams article from February:


    Humans often do not think big enough. They also don’t think small enough. How many aliens are actually out there sending radio waves from the surface of a planet into the galaxy? As opposed to so many much more advanced and efficient methods? And how many are actually bothering to send signals or probes our way on purpose?

  • Alex Tolley July 18, 2014, 15:04

    @Mark – and importantly, those dust grains should be in the geologic strata, capturing 100’s of millions of years, even billions of years of. The trick is how to recognize them. Are they dumb, or do they activate under some stimulus, and can we recognize their activation?

    I once considered writing a short story about this very scenario.

  • RobFlores July 18, 2014, 15:52

    This seems more plausible as a strategy in interstellar probe operations
    than the larger infrastructure heavy concepts. With small micro-probes that have the ability to cooperatively repair and adapt, you have the low
    mass needed to push the speeds up. With Inexpensive individual sub-units
    the design can be used for several target stars.
    As an inspiration lets remember the mircrobe Deincoccus radiodurans microbe that can survive in 500,000 rads at gamma ray energies. It should be a hint that with the right design a probe ‘cloud’ moving at significant % of C can make an interstellar journey.

  • Eniac July 19, 2014, 11:07

    It is often assumed here unquestioned that it is easier to accelerate small craft than large. Is that really true? I have my doubts. Yes, it takes less energy, but a smaller craft can carry/receive/process less energy, as well. The two factors quite possibly cancel. Add to this the above mentioned communications and the shielding problem, small craft do not look all that good for interstellar travel.

    Note also that some propulsion methods, notably magnetic sails or nozzles, do not work on a small scale, at all.

    Fission fragment might work, perhaps a thin coating of Americium to parts or all of one side of the sprite could provide an effective means of propulsion requiring no external energy or force.

  • Eniac July 19, 2014, 11:27

    Assume for a moment that you can somehow electronically control the amount or directionality of the fission fragments emitted by a thin film of, say, Curium 250. A sprite with four such patches applied to the corners of one side could be wired up with an optical sensor on the other side such that it homes in on the light of a star. Fission fragment propulsion should be good for ~1% of c, perhaps, and the half-life of Curium 250 is 6900 years. A swarm of such micro-craft could conceivably make it to Alpha Centauri in a few hundred years, at very minimal cost.

  • railmeat July 19, 2014, 21:38


    “It is often assumed here unquestioned that it is easier to accelerate small craft than large. Is that really true?”

    Yes that is really true. The smaller spacecraft is lower mass, thus easier to accelerate.

    I interpret your question to be can we accomplish what we want faster and better with a smaller spacecraft. That is not so clear. I favor smaller spacecraft because each one will be cheaper and easier to build. That means we can create new designs and try them out quicker. We will iterate to a useful solution more quickly that way. More experiments per year, or decade, mean faster learning. That is the primary advantage to smaller spacecraft.

    I hope we will see a lot more like Kicksat and professor Peck’s ideas.

  • Eniac July 20, 2014, 12:28


    Yes that is really true. The smaller spacecraft is lower mass, thus easier to accelerate.

    You are simply repeating the unquestioned assumption. Here and now, what is easier to accelerate, a bullet or a grain of sand? According to your categorical statement, a grain of sand. Yet, in reality, there is currently no way to accelerate a grain of sand to supersonic velocities, and it would not be easy to devise one.

  • David W July 20, 2014, 20:49

    Thanks for answering my Question in the last post this Way! This is very exciting work

  • railmeat July 20, 2014, 21:20


    That a smaller mass is easier to accelerate is not an unquestioned assumption. It has already been resolved by Newton.

    A smaller, lighter spacecraft will require less energy to accelerate to the velocity we need than a larger heavier one.

    I agree that there are a whole host of other engineering constraints that mean that the optimal size of a spacecraft for an interstellar mission is not necessarily the smallest possible. However at this early stage I think we would get the most benefit from experimenting with shrinking spacecraft and trying to do the most with the least mass.

  • Andrew Palfreyman July 22, 2014, 15:39

    Eniac is raising the question of the scaling of propulsion unit thrust versus the scaling of the all-up mass. It’s not a trivial question, and it certainly has different answers at different scales.