≡ Menu

Can Kepler be Revived?

Never give up on a spacecraft. That seems to be the lesson Kepler is teaching us, though it’s one we should have learned by now anyway. One outstanding example of working with what you’ve got is the Galileo mission, which had to adjust to the failure of its high-gain antenna. The spacecraft’s low-gain antenna came to the rescue, aided by data compression techniques that raised its effective data rate, and sensitivity upgrades to the listening receivers on Earth. Galileo achieved 70 percent of its science goals despite a failure that had appeared catastrophic, and much of what we’ve learned about Europa and the other Galilean satellites comes from it.


Image: Galileo at Jupiter, still functioning despite the incomplete deployment of its high gain antenna (visible on the left side of the spacecraft). The blue dots represent transmissions from Galileo’s atmospheric probe. Credit: NASA/JPL.

Can we tease more data out of Kepler? The problem has been that two of its four reaction wheels, which function like gyroscopes to keep the spacecraft precisely pointed, have failed. Kepler needs three functioning wheels to maintain its pointing accuracy because it is constantly being bathed in solar photons that can alter its orientation. But mission scientists and Ball Aerospace engineers have been trying to use that same issue — solar photons and the momentum they impart — to come up with a mission plan that can still operate.

You can see the result in the image below — be sure to click to enlarge it for readability. By changing Kepler’s orientation so that the spacecraft is nearly parallel to its orbital path around the Sun, mission controllers hope to keep the sunlight striking its solar panels symmetric across its long axis. We may not have that third reaction wheel, but we do have the possibility of this constant force acting as a surrogate. Re-orientation of the probe four times during its orbit will be necessary to keep the Sun out of its field of view. And the original field of view in the constellations of Cygnus, Lyra and Draco gives way to new regions of the sky.

Testing these methods, mission scientists have been able to collect data from a distant star field of a quality within five percent of the primary mission image standards. That’s a promising result and an ingenious use of the same photon-imparted momentum that’s used in the design of solar sails like JAXA’s IKAROS and the upcoming Sunjammer sail from NASA. We now continue with testing to find out whether this method will work not just for hours but for days and weeks.


Image (click to enlarge): This concept illustration depicts how solar pressure can be used to balance NASA’s Kepler spacecraft, keeping the telescope stable enough to continue searching for transiting planets around distant stars. Credit: NASA Ames/W Stenzel.

If a decision is made to proceed, the revised mission concept, called K2, will need to make it through the 2014 Senior Review, in which operating missions are assessed. We should expect further news by the end of 2013 even as data from the original mission continue to be analyzed.

All of this may take you back to the Mariner 10 mission to Mercury in 1973, itself plagued by problems. Like Galileo, its high-gain antenna malfunctioned early in the flight, although it would later come back to life, and like Kepler, Mariner 10 proved difficult to stabilize. Problems in its star tracker caused the spacecraft to roll, costing it critical attitude control gas as it tried to stabilize itself. The guidance and control team at the Jet Propulsion Laboratory was able to adjust the orientation of Mariner 10’s solar panels, testing various tilt angles to counter the spacecraft’s roll. It was an early demonstration of the forces at work in solar sails.

We can hope that ingenuity and judicious use of solar photons can also bring Kepler back to life in an extended mission no one would have conceived when the spacecraft was designed. What we’ll wind up with is about 4.5 viewing periods (‘campaigns’) per orbit of the Sun, each with its own field of view and the capability of studying it for approximately 83 days. As the diagram shows, the proper positioning of the spacecraft to keep sunlight balanced on its solar panels is crucial. It’s a tricky challenge but one that could provide new discoveries ahead.

Addendum, from a news release just issued by NASA:

Based on an independent science and technical review of the Kepler project’s concept for a Kepler two-wheel mission extension, Paul Hertz, NASA’s Astrophysics Division director, has decided to invite Kepler to the Senior Review for astrophysics operating missions in early 2014.

The Kepler team’s proposal, dubbed K2, demonstrated a clever and feasible methodology for accurately controlling the Kepler spacecraft at the level of precision required for scientifically valuable data collection. The team must now further validate the concept and submit a Senior Review proposal that requests the funding necessary to continue the Kepler mission, with sufficient scientific justification to make it a viable option for the use of NASA’s limited resources.

To be clear, this is not a decision to continue operating the Kepler spacecraft or to conduct a two-wheel extended mission; it is merely an opportunity to write another proposal and compete against the Astrophysics Division’s other projects for the limited funding available for astrophysics operating missions.


Comments on this entry are closed.

  • ljk December 4, 2013, 10:30

    The problems with Voyager 2’s main and backup radio receivers, the latter of which it has been using since 1978, should be more than enough inspiration and support for a continuation of Kepler.

    Voyager 2 also managed to explore two additional planets and their moons and rings despite not being part of the original mission objectives, all with a wonky backup radio receiver made in the 1970s.

    Quoting from here:


    “In April 1978, due to a ground error resulting in the failure to send expected commands to Voyager 2, the spacecraft’s fault-protection software assumed the primary radio receiver had failed and switched to the backup receiver.

    “When this happened, the backup receiver suffered a short in a key component: the tracking loop capacitor. This left the backup receiver unable to adjust for frequency drifts of the signal from Earth.

    “When controllers commanded Voyager 2 to resume using the primary receiver, the primary receiver failed completely, resulting in a week of waiting for the fault-protection software to again command use of the degraded backup receiver.

    “Controllers ultimately learned to compensate for the limitations of the backup receiver.”

  • Ron S December 4, 2013, 11:26

    Hmm…that diagram seems to beg a question.

    When “ridge-on” to the sun the spacecraft is stable on its long axis. Otherwise the spacecraft is unstable and will rotate. This implies that the spacecraft may only be quasi-stable with respect to photon pressure.

    It seems to me that either of the following must be true. There is an orientation zone near ridge-on whereby the spacecraft torques back to ridge-on (negative feedback). If there is no such zone (positive feedback) the torque is low enough at small displacements for the remaining reaction wheels to compensate for photon-induced roll on the long axis.

    Is this close to correct?

  • Paul W December 4, 2013, 12:00

    Wow. Very impressive.
    How smart can desperate people get? Whether K2 can get any additional funds from cash strapped NASA may be another matter. With our government making space science such a low priority, it’s tough going for even the most worthy projects. Best of luck to them!

    That depiction of the Galileo satellite gives me the creeps when I imagine the hugely expensive JWST having to unfurl all of its mirrors, shades, and light panels far out in space. I totally support JWST, but I like Sara Seeger’s idea of nano satellites that study only one star at a time, piggybacked to a launcher dedicated to some other mission.
    Until we can get a reliable access to space to fix our toys, we have to strive to get the best info for the least money, and the least risk.

  • Paul Gilster December 4, 2013, 13:04

    Ron S writes:

    If there is no such zone (positive feedback) the torque is low enough at small displacements for the remaining reaction wheels to compensate for photon-induced roll on the long axis.

    That’s how I read this — the remaining reaction wheels are sufficient to compensate for the roll. Other comments welcome.

  • Rob Flores December 4, 2013, 13:06

    Well, I guess it’s a thrifty attempt. And it will only have the capacity to find HZ worlds orbiting lower massed K stars and M types. I am guessing
    that the 83 day window is not linear in quality and near the end of the period not as reliable. all those Tide locked planets are only moderately exiting.
    Commercial satellites launches sometimes have room for extra mini-payloads. Is there a way to design and place an External reaction wheel on the Kepler craft in a way that compensates for the dead reaction wheel. maybe this could be done for less than two million. That would bring it within private donations fundraising teritory.

  • Holger December 4, 2013, 13:29

    Ron S:

    It seems to me that either of the following must be true. There is an orientation zone near ridge-on whereby the spacecraft torques back to ridge-on (negative feedback).
    If there is no such zone (positive feedback) the torque is low enough at small displacements for the remaining reaction wheels to compensate for photon-induced roll on the long axis.

    I think the first must be true, otherwise you’d still need three reaction wheels to maintain the spacecraft’s position in all 3 dimensions. They say the position is “stable”, not just an instable equilibrium (though the diagram seems to imply the opposite…)

    So has it been stated how long the K2 mission could run? If this works for more than 4 years, it could potentially find HZ planets (with periods very close to 365 days) around Sun-like stars as well, if they point at the same 4-5 FOV’s each year. Of course, K2 will break down if still another reaction wheel fails…

  • Michael December 4, 2013, 14:13

    The diagram seems to indicate the solar wind and not light as light would be reflected/absorbed and re-emitted in straight lines. I would have also thought it would be a balancing act between light pressure and the somewhat less forceful solar wind as well.

  • M. Gauthier December 4, 2013, 17:25

    I’m amazed by how an innocuous design decision can lead to the salvation of an expensive piece of hardware. Had the designers chosen a flat panel for the sunward side of the spacecraft I’m guessing this ingenious solution would not have been possible.

    If NASA does decide to end the mission, I wonder if they could hand it over to people closer to the level of citizen scientists and amateur astronomers. The machine is up there already; assuming a private means of communicating with it is possible, allowing public access to a “dead” telescope would cost NASA nothing. Kepler could yet generate useful science for “free” in such a situation.

  • Mephane December 5, 2013, 3:52

    I am wondering whether it would be a viable approach to design spacecraft to use this approach for orientation and stabilization in the same manner, by equipping them with multiple small solar sails that can be reoriented as needed.

  • swage December 5, 2013, 6:00

    It doesn’t really belong to this topic, however this my be relevant for the overarching topic of this site, e.g. building a spaceship.

    New Solar Cell Material Acts as a Laser As Well

    Whilst i can see no obvious use for this technology in conjunction with lightsails per se, one of its applications may be turning solar cells used for powering spacecraft into dual use devices for either power generation or thrust generation as a photonic laser thruster, thus eliminating the need for an additional propulsion solution.

    I am no expert, but i think this is a significant finding for space applications.

  • ljk February 4, 2014, 23:44

    Kepler’s second act

    Last summer, NASA announced that the Kepler spacecraft could not continue its mission to look for exoplanets because of failed reaction wheels on the spacecraft. Jeff Foust reports on how the project is trying to bring new life to the spacecraft with an alternative mission, as other spacecraft seek to follow in Kepler’s footsteps.

    Monday, February 3, 2014


  • Dave C February 9, 2014, 14:29

    Finally found the fields of view for the proposed Kepler K2 mission:

    It looks like field 7 covers the area of the Wow Signal:

    Wow Signal location:

    Too bad we don’t have a star associated with the signal.
    Serendipity in action!

  • ljk February 10, 2014, 11:53

    Kepler sees its first exoplanet in months:


    To quote:

    WASP-28b is about the size of Jupiter and is in a very tight orbit around its star, with a year that lasts just 3.4 Earth days. Unfortunately, K2 will not be able to carry on with Kepler’s original quest to find habitable Earth-sized planets around sun-like stars. To confirm that a planet is real, Kepler needed to see it transit three times, meaning true Earth twins would take about three years to confirm. The modified space telescope won’t be able to maintain its lock on a star for that long.

    But the K2 mission will be able to collect data on very young stars and search for planets or planet-forming discs around them. “This will be a window into both star formation and planet formation,” says Howell.

    For now K2 is operating on funding reserves from the main Kepler mission, and the team is waiting on approval for NASA funding for another two years. That decision is expected by the end of May. In the meantime, the team is ploughing ahead. The first official science campaign will start on 1 March – and the team has already received requests to observe 110,000 target stars.

    “For us that’s great news, not because we can do all of that, but because it’s a sign of the interest of the community,” says Howell.

  • ljk February 10, 2014, 11:55

    Dave C said on February 9, 2014 at 14:29:

    “It looks like field 7 covers the area of the Wow Signal:


    “Too bad we don’t have a star associated with the signal.”

    I see we are still stuck on the paradigm that an alien signal must come from the surface a planet, their home world no less. This is why SETI has not progressed as far as it should have fifty-plus years after Ozma.

  • ljk May 16, 2014, 13:19

    Kepler Mission Manager Update: K2 Has Been Approved!

    May 16, 2014

    The team received good news from NASA HQ — the K2 mission, the two-wheel operation mode of the Kepler spacecraft observing in the ecliptic, has been approved based on a recommendation from the agency’s 2014 Senior Review of its operating missions.

    The approval provides two years of funding for the K2 mission to continue exoplanet discovery, and introduces new scientific observation opportunities to observe notable star clusters, young and old stars, active galaxies and supernovae.

    The 2014 Senior Review report is available at:


    Full news item here: