In this great answer I learned that the Mars rover Curiosity can be given some tasks and it will go ahead and manage the work and navigation by itself, to at least a certain limit.

The James Webb Space Telescope needs to maintain its attitude so that the giant mirror and other optics and experiments stay dark and cold - about 50 Kelvin, while its solar panels get enough sunlight to produce about 2 kilowatts of electricity.

It will also be in a somewhat unstable halo orbit around Sun-Earth L2. Typically spacecraft are monitored from Earth stations, and doppler measurements are precise enough to decide when orbit correcting thrust maneuvers are necessary to keep it from wandering away very much. Without careful, regular small corrections, much larger fuel-gobbling corrections would be needed.

These corrections are usual surprisingly small, of the order of a kilometer, and deep sub-meter-per-second in velocity. Precise doppler measurements are combined with detailed orbit modeling including perturbations are used to calculate each maneuver on earth, then checked and double-checked before the instructions are sent to a spacecraft in a halo orbit. Afterwards a second set of measurements are made to confirm that the maneuver had the intended correcting effect on the orbit.

Could the JWST manage its own regular orbit correction maneuvers without help from Earth if absolutely necessary? Suppose there's a failure or a meteorite hit that caused a communications problem - can it at least maintain an orbit until a repair mission (crewed or automated) comes to fix it?

If so, how can it measure its own position precisely enough so that the regular corrections are small? While star cams can be used to ascertain attitude, how will position be measured precisely? Are occasional GPS measurements possible at that distance? Does it have good enough visible light cameras to make super-precision terminator measurements or even pattern recognition of landscape features or occultation timings of/by the Earth and the Moon?

  • 7
    $\begingroup$ No. See space.stackexchange.com/questions/5845/… $\endgroup$
    – Hobbes
    Oct 5, 2016 at 14:29
  • $\begingroup$ @Hobbes since the JWST's halo orbit is a lot closer to the Earth and Moon than the interplanetary spacecraft discussed there, there are additional options in this case that would not be available at much larger distances. GPS and optical were the two I could think of off the top of my head, but there could be more. So the fact that interplanetary spacecraft can't and/or don't do it does not a priori prove that a 21st century state-of-the-art observatory in a halo orbit around Earth's SE-L2 could not. $\endgroup$
    – uhoh
    Oct 5, 2016 at 14:46
  • 2
    $\begingroup$ One important aspect is that an emergency maintain-orbit-pending repair capacity is not needed simply because no repair is likely to be sent. JWST is not designed for servicing, and even if it were, no craft which could support such a mission are likely to be in service for some time. Easier to build and launch a new one, in those circumstances... $\endgroup$
    – Andrew
    Oct 5, 2016 at 17:36
  • $\begingroup$ @Andrew robotic in-space repair is coming... someday at least. See info in this currently UNANSWERED question about in-space repair. Things that have not been designed for servicing may not be necessarily absolutely unserviceable, but I suppose it depends a great deal on what in particular needs fixing. But as much as I want JWST to have a magic self-orbit-maintaining capability because I think it would be cool, I have a feeling your logic is inescapable in this case. $\endgroup$
    – uhoh
    Oct 6, 2016 at 0:54

2 Answers 2


I don't have too much time to research as I'm about to head off to work, but a quick investigation of the JWST doesn't show anything that could be used for Astrogation.

Curiosity can do its own pathfinding in part because of the programming genius of its creators, and also because of how easily detectable the obstacles are. A camera can easily detect a hazardous boulder a few meters in front of you because it's a large, obvious object. After that detection, everything else is an exercise in pathfinding.

Astrogation is a lot more difficult, because you're pathfinding through empty space and not a boulder field. Unlike rocks and valleys, Lagrange points aren't directly detectable. You have to infer the location of a Lagrange point from other objects, and without any reference points there's nothing you can do.

Now, from what I've gleaned the JWST doesn't have any Earth-facing telescopes or rangefinders it could use to ascertain its own position. Without position data from Earth, there'd be no way for it to figure out where the Lagrange point is relative to its own location, and stationkeeping would be impossible. In a sense, Earth is more of a Beacon or Reference-Frame than it is command-and-control.

  • 1
    $\begingroup$ I suppose the giant heat shield blocking out the Sun and Earth has something to do with the lack of earth-facing telescopes or range finders :) There must be some fairly high gain antennae pointing that way for comms, but a camera with enough precision for this kind of job would probably be hampered by being stuck out on the end of a long support rather than the body of the spacecraft. What about GPS? Like radar, it will tend to give precision only along the earth-spacecraft axis, but that's OK if you get a sufficient number of readings over an extended period of time. $\endgroup$
    – uhoh
    Oct 5, 2016 at 13:59
  • $\begingroup$ Actually I think that's probably not right about the GPS. There is probably no useful triangulation using GPS in the conventional way.The only time there could possibly be signal is when the satellite is just beyond the Earth's terminator and the side lobes project past Earth. That gives you only about a quarter of 1 degree. See for example the slides and links in this UNANSWERED question $\endgroup$
    – uhoh
    Oct 5, 2016 at 15:21
  • 1
    $\begingroup$ I agree with @UIDAlexD, I didn't find anything either. $\endgroup$
    – Hobbes
    Oct 6, 2016 at 11:25
  • $\begingroup$ @Hobbes OK I guess half of me is surprised it doesn't have some kind of auto-pilot, but the other half is brought back to Earth by UIDAlexD's logic. Not this one, maybe the next one - or the one after that, when some deeper space atomic clocks get out there. $\endgroup$
    – uhoh
    Oct 8, 2016 at 7:21
  • $\begingroup$ @uhoh: I can get an earth vector with a precision of 1/4 of 1 degree every few tens of minutes? So much easier. I'm pretty sure if its autostation-keeping could keep it with in one degree (from earth's perspective) of correct that's a big win. $\endgroup$
    – Joshua
    Dec 28, 2021 at 2:59

Short answer is no--JWST does not manage its orbit corrections itself. Stationkeeping is done from the ground. During normal ops ,there are both medium-gain and high-gain antennas on the sun side, which can be used to receive commands and transmit telemetry and ranging/doppler information. This ranging and doppler is sufficient to manage orbit corrections from the ground. Like stated, these are likely to be small velocity changes whose effects will add up over time.

At L2, JWST's control volume is absolutely huge. Anywhere within said control volume is allowable for normal operations. The Spacecraft is uploaded with an Ephemeris/Quaternions, and is equipped with star trackers--so it has limited positional information, mostly for locating the Earth.

Regarding "Curiosity can be given some tasks and it will go ahead and manage the work and navigation by itself, to at least a certain limit", JWST is also observation-driven. Each "Visit" that is uploaded contains a list of observations to take, and instructs the observatory to slew all over the place. It will slewing from A to B to C... during a Visit, with occasional breaks for data transmission and autonomous momentum unloads. So in this sense, slewing to a new attitude, acquiring and guiding on chosen stars, etc, is similar to what Curiosity does. Except JWST does not have move along the ground, only in 3-D space.

By using the DSN, an "accident" on the ground is unlikely to have long-standing effects with JWST communications without impacting the prime and redundant control centers (STScI/GSFC). Should this happen for a long enough period, the spacecraft would react to a loss-of-comm, and reconfigure and reorient itself to maximize power and chance of communications--it could sit this way indefinitely.

  • 1
    $\begingroup$ This is a really interesting answer, thank you! Is it possible to clarify a few things; As far as "it could sit this way indefinitely", this answer links to this longer answer which explains that even in a pure CR3BP scenario only the smallest halo orbits are stable, (not one as large as JWST's). Isn't JWST's orbit exponentially unstable? and aren't the bi-monthly station-keeping corrections absolutely critical to conserving fuel and keeping JWST in its orbit for it's mission duration? $\endgroup$
    – uhoh
    Jul 24, 2019 at 23:24
  • $\begingroup$ more about the necessity of JWST's two week station-keeping cadence and delta-v budget in these answers: 1, 2, 3. Without this regular station-keeping doesn't it end up drifting away exponentially along one of the unstable manifolds? $\endgroup$
    – uhoh
    Jul 24, 2019 at 23:33
  • $\begingroup$ @uhoh This answer is spot-on. The JWST does not navigate its translational state. It doesn't need to. With regard to "accidents", they happen with great regularity. Another name for those "accidents" is weather. The Ka band used to downlink reordered imagery data are highly susceptible to precipitation. One of the three locations of the DSN ground stations (Goldstone) is in the western US high desert. The other two (Madrid and Canberra) are in semi-dry climates. $\endgroup$ Jan 27, 2022 at 12:57
  • 1
    $\begingroup$ While the Ka band data used for imagery data (downlink) might not get through due to clouds and precipitation, the S band used for housekeeping telemetry (downlink) commanding (uplink), and ranging (uplink and downlink) is much less susceptible to weather. "Accidents" (aka weather) might (and probably will) result in loss of some imagery data. The range and range rate data are much less likely to suffer those whether-based accidents. $\endgroup$ Jan 27, 2022 at 13:01
  • $\begingroup$ @uhoh The short answer is "no" is correct. The JWST cannot manage its own orbit if necessary. It's that simple. $\endgroup$ Jan 27, 2022 at 19:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.