NASA Airbag Lander Technology - 100% success rate so far

Hi,

With the Opportunity Rover landing today, NASA (JPL) has successfully
landed 3 spacecraft on Mars using airbag technology, that makes it a
100% success rate.

Do you think the floodgates will open for future Mars landers using
airbag technology on more interesting locales (near mesas, volcanoes,
valleys, etc)

Mark Rejhon

In article ,
Mark Rejhon wrote:
With the Opportunity Rover landing today, NASA (JPL) has successfully
landed 3 spacecraft on Mars using airbag technology, that makes it a
100% success rate.

Which statistically is indistinguishable from the success rate of rocket
landing. (If *one event* happening differently would change the relative
ranking of the options, you don't have enough data for that ranking to be
trustworthy.)

Do you think the floodgates will open for future Mars landers using
airbag technology on more interesting locales (near mesas, volcanoes,
valleys, etc)

Unlikely. The airbag/rocket landing method (yes, there are rockets
involved, and a radar altimeter too) remains complex, heavy, poorly
controlled, and hard on the payload (MP's first bounce was 18G). And
even in the days before it acquired braking rockets, the expectation
was a landing failure rate of perhaps 10%.
--
MOST launched 30 June; science observations running | Henry Spencer
since Oct; first surprises seen; papers pending. |

(Mark Rejhon) wrote in
om:

Hi,

With the Opportunity Rover landing today, NASA (JPL) has successfully
landed 3 spacecraft on Mars using airbag technology, that makes it a
100% success rate.

Do you think the floodgates will open for future Mars landers using
airbag technology on more interesting locales (near mesas, volcanoes,
valleys, etc)

More 'interesting' (read: risky) sites will probably require active
landing aids to avoid pitfalls, or precise navigation to verified
safe locations; that might point to powered landings.

How far up in payload/size can airbags be scaled?

Just the same, airbags have demonstrated repeated success and
remain a relatively simple/less expensive method of landing. I
expect that technology will be used frequently.

--Damon

(Mark Rejhon) writes:

With the Opportunity Rover landing today, NASA (JPL) has
successfully landed 3 spacecraft on Mars using airbag technology,
that makes it a 100% success rate.

A sample of 3 does not make any useful statistics.

Do you think the floodgates will open for future Mars landers
using airbag technology on more interesting locales (near mesas,
volcanoes, valleys, etc)

Foodgates? No.
Mars missions will remain at a trickle for the foreseeable future.

If you want to land something bigger then the current rovers you'll
probably have to go for rocket powered descent because airbag systems
don't scale that well. Going down into valleys would reduce available
solar power and limit communications, so you won't see that for a
while. Maybe once we have a network of Mars comms satellites and
nuclear powered rovers...

--
Manfred Bartz

(Henry Spencer) wrote in message ...
Which statistically is indistinguishable from the success rate of rocket
landing. (If *one event* happening differently would change the relative
ranking of the options, you don't have enough data for that ranking to be
trustworthy.)

Nice to hear from you - Henry Spencer

Since you know far more than I do about space history, you
might be right. I know retrorockets has their own advantages,
but don't you think airbags are making a useful addition at least --
One can really get tempted to look at the history of Mars
landers (Ones with no airbag shock absorbers).
There seems to be a very dismal success rate with landers,
considering that even Mars Polar Lander failed.

But since the Beagle 2 failed (another airbag lander),
that would mean a universal success rate of 3 out of 4 so far.
That's a definite strike.

Do you think the floodgates will open for future Mars landers using
airbag technology on more interesting locales (near mesas, volcanoes,
valleys, etc)

Unlikely. The airbag/rocket landing method (yes, there are rockets
involved, and a radar altimeter too) remains complex, heavy, poorly
controlled, and hard on the payload (MP's first bounce was 18G).

I'm familiar with the rockets and radar altimeter; I'm reasonably
technically literate to know (google for my last name Rejhon
to see).

However, since you know more about the underlying technologies
in the history of space, it may just be that airbag landers
is still too unproven a technology.

For the purpose of this post, let's define my terminology
"airbag lander" as any spacecraft lander that uses airbag
as a shock absorber at surface contact, even if additional
sethods of slowdown is employed (retrorockets, like those
used for Spirit and Opportunity)

I do remember that Pathfinder landed really hard,
but Spirit was relatively gentle despite worse air conditions
including clear-air-turbulence and massive swinging of the lander.
The post-landing video computer recreation of Spirit's airbag
landing showed what looks like a very dramatic swing
swing just a few feet above the ground (I don't remember how
extreme, but it was almost 30 degrees if I remember correctly!)
with lots of horizontal velocity successfully cancelled out by
the retrorockets before the the lander was released.
I was impressed. I wonder if retrorockets would have
survived the clear-air turbulence that occured with Spirit,
unless they were well designed retrorockets, while being
reasonably economical in a lander? Definitely, the airbag
method now sounds sounds cheaper and safer that way --
as long as airbag is viewed simply as a substitute
to landing-leg shock absorbers. Much more forgiving
compared to a landing-legs lander accidentally tipping
over (the airbag lander just rolls until it is stable...)

Now, I heard that Opportunity's lander was amazingly gentle
when I was watching the realtime status reports that
occured at http://spaceflightnow.com/mars/mera/statustextonly.html
it was only about 2-3 G's! That seems gentler than the vast
majority of landing-leg landers.

Only 2-3 G's force for the first bounce of the
Opportunity airbag lander -- very gentle.

The argument rises in that airbag landers are more forgiving
of terrain. One of the Viking landers (Viking 2) landed with
one of its legs sitting on a rock. It retrorockets did not
shut down until a little late, because of faulty altimeter data,
and the retrorockets blew up more dust than predicted
(more than the 1mm of dust that NASA predicted).
The retrorockets disturbed the landing site. An airbag lander
would have handled that particular situation better, and would
probably work with larger rocks. (Let's compare apples to apples,
Lander vs Lander, not Lander vs Rover of course -- tilt might
be too extreme for a rover if an airbag lander . Besides,
Spirit and Opportunity are designed to be able to conduct a
partial mission in its undeployed lander state even at extreme
tilt). If only two out of three petals are opened, some
science can still be done! Like when the airbag lander
lands against the bottom of a huge bolder and one petal
is jammed against the boulder, even if the base petal is
the one jammed against the boulder. Some view would be
blocked and pancam will not be possible to deploy, but some
photos from at least some of the cameras would still be taken
and sent back to Earth by the low-gain antenna. As well
as other science information such as atmosphere, temperature,
etc. An upsidedown traditional landing-leg lander would
probably now be as dead as an upsidedown turtle unless
it was designed to resist a tip over or execute
tip-over recovery (i.e. robotic arm strong enough to
push itself up)

Another scenario arises with the crater that Opportunity
landed inside of. What if a traditional landing-leg lander
(i.e. one that uses shock absorbers in landing legs only)
lands on a steep slope of a crater - especially if
there was some pretty steep elements that causes a lander
to tip over? An airbag lander would just simply roll to
the bottom of the crater. The G-forces of rolling down
a steep hill will not be big considering the low gravity
of Mars, once the shock of the first few bounce(s) have
been absorbed.

Perhaps the best of all worlds is that airbags are simply
used as a substitute to landing-leg shock absorbers,
to gain the following benefits:
- Resist horizontal velocity better than traditional landing-legs landers
Although airbags aren't recommended for horizontal movement,
they are more forgiving than landing-leg landers here;
- Land at any angle (upside down, etc)
- Roll away from steep slopes like the slope of a crater
- Roll off a large rock (hopefully with enough bounce to go
far away enough to deploy all 3 petals)
- Resist a fall off a small cliff (at low gravity, this is
theoretically safer than bouncing against sharp rocks)
- What seems to be potentially better margin of G-force
safety than landing-legs (at least in the limited data
set of the Spirit and Opportunity rovers)

It is very tempting to think that the floodgates will now open
for airbag landers in the next 10-15 years at least for miniature
probes, now that NASA seems to be quickly maturing the technology
and finally adding an order of margin safety (Opportunity Rover
designed to resist 40G, but endured only little more than 2G at
first bounce!)

I do agree that more things have to run correctly during
re-entry for airbag landers, but when you try to design
a small lander at tighter budgets than Viking, the airbag
design seems safer. (Witness Pathfinder/Spirit/Opportunity
versus Polar Lander). There are ocasions in history
where actual landing-leg landers got killed by
turbulence, tipover, horizontal movement, and more.
(Just look at all the soviet lander attempts, at least
based on the interpretation of the telemetry of these
attempts). Airbag landers can survive moderates amounts
of all of the above much better than landing-leg landers.
Despite more things that can go wrong with the lander
technology, the final-step (surface contact) is still
the riskiest and landing-leg landers have fared very
poorly historically on Mars (Remember: Mars has
atmosphere and weather -- let's not compare the
highly successful moon landings here.).

I am seeing how quickly airbag landers are being matured
versus landing-leg landers (and even recent ones at that).
Lower G forces (Pathfinder - Spirit - Opportunity)
and resistance to massive air turbulence (Spirit).

Retorocket landers definitely have their advantages -
but why not simply add airbags as a tipover safety measure?
Big landers such as a human lander, a large nuclear-powered
rover, or a rocket fuel factory, probably will have to be
landed in the traditional retrorocket manner. Then again,
maybe airbags might be applied to large landing-leg landers at
least as a basic safety measure - like airbags in a car.
So if a human lander tips over unexpectedly, compact
concealed airbags might save the lives of the astronauts and
save equipment damage and they would be able to exit the
lander (through a small non-spacelock emergency exit in
spacesuits) and winch the lander back to its upright position.
It would only be designed to absorb a tipover shock, and
only deploy if the lander angle is too extreme. Just small
airbags good enough to prevent mission-ending damage on
lander tipover. Might even be smaller airbags (but
in huger numbers) than the ones used to land the rovers.

(Sorry about grammar/spelling errors, I typed this in
one pass in Google Groups)

Mark Rejhon
http://www.marky.com

Part #2
Addendum to things I did not fully respond.
(I'm too lazy to proofread on things I missed.)


Statistically indistinguishable

Statistically if we're aiming for a good large
set of statistics like "95% accurate 19 out of 20 times".

But the statistics are starting to be significant
enough to be worth discussing even if not industry-standard
accurate like above yet.

Even flipping a coin three times to land on one side is
still a 1 in 8 chance. NASA succeeded in doing that.
Success rate of 3 out of 3. If we're talking about a
pessimistic 50% success rate. I can argue that
landing-leg landers have less than a 50% success
rate on an atmospheric planets other than Earth
with unknown and unpredictable weather.
This explicitly excludes Mars and asteroids,
so the odds are worse than 1 in 8 to succeed
3 consecutive attempts.

Now, if you view the statistics that way, that starts
to become vaguely statistically significant, an event
meriting watching. Not ironclad statistics, but
I argue that it *is* reasonably statistically
significant now. :-)

It's not like airbags are an unproven technology;
especially with their proven life-saving statistics
with cars. They only needed to make it more robust
for space use. That's a lot easier than designing
a retrorocket lander, and easier to quickly advance
this art. Early successes with airbags are
vastly better than the early successes with retorockets
(and still, even better than the last few retorocket
landers on atmospheric planets by any nation).
The difference is so dramatic that it's already
at least somewhat statistically significant after just
a few attempts now.


(If *one event* happening differently would change the
relative ranking of the options, you don't have enough
data for that ranking to be trustworthy.)

Good point, but let's pretend we're now using airbags instead of
landing legs. Design the mission to be as similiar as possible
to landing leg wherever possible, with the sole exception of a
parachute cable that needs to be released. Try to do a
traditional landing-leg mission in direct-to-Mars trajectory
versus airbag-shock-absorber mission in direct-to-Mars trajectory,
and the design goals end up becoming surprisingly similiar
especially when you read through the lines. A supersonic
parachute is needed in both cases. Pre-re-entry retrorockets
apply in both cases. Some of the most dangerous weak links
are surprisingly identical -- a supersonic parachute needed
for a direct-to-surface mission to Mars without orbiting first
like Viking.

When one views it this way, let's finally concentrate on
the main difference in final shock absorbing.
Airbags are more resistant to problems like tipover,
horizontal velocity, one event going wrong.
Airbags tend to deploy more reliably than retrorockets
can fire.

Also, generally, airbag statistics of all airbag projects in
the last 50 years combined (including cars), versus all
rocket projects (including non-space rocketry) or
even just space rocketry, you see that statistics are
much better with reliable deployment of airbags!
So the few extra steps (airbag inflation, etc)
were pretty reliable technology even at the time
Pathfinder was launched. They just needed to make
the airbags stronger due to sharp rocks, that was
probably the biggest difficulty.

Also, if retrorockets in an airbag landers failed -- this does
not guarantee the end of an airbag lander mission; it's still
possible for an airbag lander to survive without retrorockets
(complete failure for final retrorockets).
G forces may hit or slightly exceed critical limits.
But that means the lander *may* have survived.
Landing leg landers are less likely to survive at the speed that
an airbag lander hits the surface, if there are 100% failure
of landing retrorockets in either cases. Unless you spend
lots of money reinforcing the lander to survive a parachute-speed
impact. Even that doesn't have a good success rate historically
either (even in the limited data). That's very tricky on Mars
with a thin atmosphere that's not very good for parachute-only
landings.

Also, even if the parachute cable is not released, that doesn't
guarantee the end of an airbag lander either. So I consider this
so-called critical step as "semi-critical". More often
than not (more than 50% of the time), parachutes will float
sideways and land next to the lander, still permitting deployment.

Once you think of it this way, it's pretty easy to suspect
that the so-called "15 critical steps" is really not much worse
off than a theoretical similiarly-designed
direct-to-surface-of-atmospheric-planet landing-leg lander!
Plus, more than compensated by the better airbag resistance to
a bad velocity vector at actual surface contact.
(Better resistance to surface contact speed at
any direction and orientiation than a landing-leg
lander, including unexpected events caused by weather)

Airbags are no longer necessarily hard on payload anymore;
as shown by Spirit and Opportunity. Worse has already
happened to landing-leg landers, many have failed because
they "crashed" (due to any failure in the landing steps,
whether horizontal velocity, tipover, extreme velocity,
premature shutoff or failure of retrorockets, etc).

NASA has added much better G-force margins already since
Pathfinder anyway with the relatively gentle Spirit landing
and the extremely gentle Opportunity landing.
Airbag landers should have a much better chance of
surviving all that, more than compensating for the
complex steps of a lander mission (even the lander
missions that might simplify risk in certain ways by
orbiting before landing, ala Viking. But that is
expensive to do.)

Mark Rejhon
http://www.marky.com

Addendum:

This explicitly excludes Mars and asteroids,

Ooops.
I mean this explicitly excludes Moon and asteroids. :-)

Again, too lazy to proofread. Whoops!
But I bet you figured that out already.

Mark Rejhon
http://www.marky.com

In article ,
Damon Hill wrote:

More 'interesting' (read: risky) sites will probably require active
landing aids to avoid pitfalls, or precise navigation to verified
safe locations; that might point to powered landings.


Why not use a ballute to slow down the probe and give it some "hover
time". A small engine could give the ballute and probe some cross range,
and would allow for a very soft landing near interesting terrain (say,
near the base of one of the canyon walls that show signs of very recent
fluid flow). The probe/rover would also have to be nuclear powered, to
extend its surface life.

There was also a Russian/Planetary society design that involved a hot
(Martian) air balloon that dangled a small probe onto the surface at
night, but the paylod was tiny.

Frank

--
Frank Henriquez Programmer/Analyst Jules Stein Eye Institute, UCLA
http://www.bol.ucla.edu/~frank/index.htm

Manfred Bartz wrote in message ...
(Mark Rejhon) writes:

With the Opportunity Rover landing today, NASA (JPL) has
successfully landed 3 spacecraft on Mars using airbag technology,
that makes it a 100% success rate.

A sample of 3 does not make any useful statistics.

Do you think the floodgates will open for future Mars landers
using airbag technology on more interesting locales (near mesas,
volcanoes, valleys, etc)

Foodgates? No.
Mars missions will remain at a trickle for the foreseeable future.

If you want to land something bigger then the current rovers you'll
probably have to go for rocket powered descent because airbag systems
don't scale that well. Going down into valleys would reduce available
solar power and limit communications, so you won't see that for a
while. Maybe once we have a network of Mars comms satellites and
nuclear powered rovers...


I am not convinced that the airbag landing system is succesful. Both
of our mars rovers are suffering from malfunctions. I don't know if
the problems are due to a hard landing or multiple bounces or both.

When the Vikings landed they landed with rockets and they functioned
after landing. If I remember right.

Zoltan

Damon Hill wrote:
[...]
More 'interesting' (read: risky) sites will probably require active
landing aids to avoid pitfalls, or precise navigation to verified
safe locations; that might point to powered landings.

Hmmm, this brings to mind idea for providing landing aids: A very
simple probe consisting of a lawn dart with an "insect bot" payload.

The lawn dart would have a significant crumple zone in the nose to
protect the bot, and a transmitter suitable for communicating with an
orbiter. There would still be a need for chutes.

The bot would have the relatively simple job of mapping the boulder
field around the lawn dart with very basic instrumentation (e.g., bump
switch and step counter). It might take several lawn darts and bots
to map a landing field large enough for a real lander and payload.
They might need to use cooperative behaviour, but I think the research
is already far enough along for the simple task.

Such lawn darts would not be a dedicated mission, but a piggyback on
an orbiter (a la Beagle 2 and MEx, and the polar landers). They would
be used to scout a prospective site for a following lander, which
could use the darts' transmitters for final approach guidance.

How does this compare to probe that is able to resolve radar/optical
images during final approach to avoid small boulders in an Armstrong
maneuver?

/dps