Bennu’s Boulders Shine as Beacons for NASA’s OSIRIS-REx


This summer, the OSIRIS-REx spacecraft will undertake NASA’s first-ever attempt to touch the surface of an asteroid, collect a sample of it, and safely back away. But since arriving at asteroid Bennu over a year ago, the mission team has been tackling an unexpected challenge: how to accomplish this feat at an asteroid whose surface is blanketed in building-sized boulders.

Using these hazardous boulders as signposts, the mission team developed a new precision navigation method to overcome the challenge.

In late August, the OSIRIS-REx spacecraft will navigate to asteroid Bennu’s surface for its first sample collection attempt. To do this, it will use an onboard image software known as Natural Feature Tracking (NFT) — a form of optical navigation that is completely autonomous.

Credits: NASA’s Goddard Space Flight Center/Scientific Visualization Studio

This video is in the public domain and can be downloaded from the Scientific Visualization Studio.

The OSIRIS-REx team had originally planned to use a LIDAR system to navigate to Bennu’s surface during the Touch-And-Go (TAG) sample collection event. LIDAR is similar to radar, but it uses laser pulses rather than radio waves to measure distance. The OSIRIS-REx Guidance, Navigation, and Control (GNC) LIDAR is designed to navigate the spacecraft to a relatively hazard-free surface. The mission had originally envisioned a touchdown site 164 ft (50 meters) in diameter, but the largest safe areas on Bennu are much smaller. The biggest site is just 52 ft (16 m) wide, or roughly 10% of the safe area envisioned. The team realized that they needed a more precise navigation technique that would allow the spacecraft to accurately target very small sites while dodging potential hazards.

In the face of this challenge, the OSIRIS-REx team switched to a new navigation method called Natural Feature Tracking (NFT). NFT provides more extensive navigation capabilities than LIDAR, and is key for executing what the team is calling “Bullseye TAG,” which delivers the spacecraft to the much smaller sampling area. As an optical navigation technique, it requires the creation of a high-resolution image catalog onboard the spacecraft.

Earlier this year, the spacecraft made reconnaissance passes over the mission’s primary and backup sample collection sites, designated Nightingale and Osprey, flying as close as 0.4 miles (625 m) over the surface. During these flyovers, the spacecraft collected images from different angles and lighting conditions to complete the NFT image catalog. The team uses this catalog to identify boulders and craters unique to the sample site region, and will upload this information to the spacecraft before the sample collection event. NFT autonomously guides the spacecraft to Bennu’s surface by comparing the onboard image catalog with the real-time navigation images taken during descent. As the spacecraft descends to the surface, NFT updates its predicted point of contact depending on the spacecraft’s position in relation to the landmarks.

graphic showing satellite, asteroid and pictures of boulders on asteroid

During the sample collection event, Natural Feature Tracking (NFT) will guide NASA’s OSIRIS-REx spacecraft to asteroid Bennu’s surface. The spacecraft takes real-time images of the asteroid’s surface features as it descends, and then compares these images with an onboard image catalog. The spacecraft then uses these geographical markers to orient itself and accurately target the touchdown site.

Credits: NASA/Goddard/University of Arizona

On the ground, team members created “hazard maps” for both the Nightingale and Osprey sites to document all of the surface features that could potentially harm the spacecraft, like large rocks or steep slopes. The team used the image catalog in conjunction with data from the OSIRIS-REx Laser Altimeter (OLA) to create 3D maps that closely model Bennu’s topography. As part of NFT, these maps document boulder heights and crater depths, and guide the spacecraft away from potential hazards while targeting a very small site. During descent, if the spacecraft predicts it will touch unsafe terrain, it will autonomously wave-off and back away from the surface. However, if it sees that the area is free of hazards, it will continue to descend and attempt to collect a sample.

NFT will be used in April to navigate the spacecraft during its first sample collection rehearsal. The operations team performed preliminary testing during the Orbital B mission phase in late 2019, and the results demonstrated that NFT works in real-life conditions as designed. NFT will also be used for navigation during the second rehearsal planned for June.

OSIRIS-REx’s first sample collection attempt is scheduled for late August. The spacecraft will depart Bennu in 2021 and is scheduled to deliver the sample to Earth in September 2023.

NASA’s Goddard Space Flight Center in Greenbelt, Maryland, provides overall mission management, systems engineering, and the safety and mission assurance for OSIRIS-REx. Dante Lauretta of the University of Arizona, Tucson, is the principal investigator, and the University of Arizona also leads the science team and the mission’s science observation planning and data processing. Lockheed Martin Space in Denver built the spacecraft and provides flight operations. Goddard and KinetX Aerospace are responsible for navigating the OSIRIS-REx spacecraft. OSIRIS-REx is the third mission in NASA’s New Frontiers Program, which is managed by NASA’s Marshall Space Flight Center in Huntsville, Alabama, for the agency’s Science Mission Directorate in Washington.

For more information on NASA’s OSIRIS-REx mission, visit:


by Brittany Enos

University of Arizona

Last Updated: March 9, 2020

Editor: Karl Hille

0 0 vote
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
March 10, 2020 2:40 am

If the asteroid is spinning or rotating, it will be much harder to “land” to collect any samples. . . just sayin…


Mark Broderick
Reply to  Jon P Peterson
March 10, 2020 3:04 am

The Moon “spins”, the Earth “spins” and we have no problem landing on them….

Len Werner
Reply to  Jon P Peterson
March 10, 2020 6:14 am

Spinning with respect to what? Earth is spinning, but hundreds of thousands of aircraft–and millions of birds, and billions of insects–land on it every day, with precision. Once the frame of reference becomes the asteroid, there no longer is any ‘spin’.

Don K
Reply to  Len Werner
March 10, 2020 8:39 am

“Once the frame of reference becomes the asteroid, there no longer is any ‘spin’.”

Yes … but … The problem is getting into the asteroid’s frame of reference. I expect it’s nowhere near the problem it is with Earth where a low altitude satellite has to lose something like 8000 mps of in track velocity and pick up a latitude dependent rotational velocity (relative to the Earth) of as much 500 mps and avoid arriving at the Earth’s surface with any significant vertical velocity.

But it’s still a problem. It looks to be a solvable problem. But I’m glad it’s not my problem.

Reply to  Len Werner
March 10, 2020 9:26 am

As Don K mentioned its not so easy to “match spin” because as you get closer the rotational velocity changes and you need to slow down, but to further complicate equations, your center of rotation is not necessarily the center of the asteroid, and the asteroid is rotating about its own center. Not so easy a calculation that needs to be autonomously used in an even more complicated algorithm to determine the appropriate thrust burns and orientations.
BTW the fuel supply is limited.

Crispin in Waterloo
Reply to  Rocketscientist
March 10, 2020 7:53 pm

Worse, it rotates about its centre of gravity, not it centre of diameter. If it is rotating rapidly, the spacecraft will be sent outwards by centripetal acceleration. Will it have to burn fuel continuously to hold the probe against the surface?

Don K
Reply to  Rocketscientist
March 11, 2020 6:52 am

“Your center of rotation is not necessarily the center of the asteroid.”

Indeed. I overlooked that. Fortunately, Bennu rotates fairly slowly. The period is about 4.3 hr. So even at the “equator” its rotational velocity looks to be only about 20cm/sec. That’s less than 1kph. The rock is roughly spherical so the vertical velocity (“bounce”) of the surface will probably be significantly less than that. That’s pretty slow. “Docking” to Bennu probably wouldn’t be that much of a problem. … If there were anything to grab hold of. But there isn’t? So velocities have to match exactly?

Bryan A
Reply to  Jon P Peterson
March 10, 2020 12:30 pm

All they should need to do is to Bump the surface with a collecting scoop that, if designed properly, Grabs and closes a scoop of surface material and draws it back into the Probe as it bounces back off the surface. Provided there are no large rocks in the scoop holding it open as it tries to close, the sample should be gathered. No need to land at all, just Bump Grab and Bounce.
Now, how is the sample gatherer designed to function??

Greg Cavanagh
Reply to  Bryan A
March 10, 2020 2:13 pm

Didn’t they use a projectile and collect the jettisoned dust some years ago?

Or as you say, a launched javelin style collector and a quick retrieval.

Reply to  Bryan A
March 10, 2020 4:33 pm

If the “bump” contact includes significant lateral components due to rotation or glancing off an icy rock the satellite could spin into the asteroid and all will be lost.
Think of drifting towards a big jagged iceberg that is spinning and creating an eddy and all you have to fend-off with is a long soda straw. What could go wrong?

Bryan A
Reply to  Rocketscientist
March 10, 2020 8:46 pm

So then the Bump would need to be as perpendicular as possible

Andy Espersen
March 10, 2020 2:57 am

Yes, Jon P Peterson – and there must also be a minute amount of gravity from the asteroid that will continuously be a problem. Exciting.

March 10, 2020 3:20 am

Yes and taxpayers would be spinning if they knew the cost of this monumental missdirection of spending!

shortus cynicus
Reply to  Warren
March 10, 2020 4:48 am

It is founded by selling t-shirts to people who need to compensate for lack of something.
According to Einstein, the demand is unlimited.

Tom in Florida
Reply to  Warren
March 10, 2020 4:50 am

Says King John II of Portugal.

Steve Taylor
Reply to  Warren
March 10, 2020 7:27 am

Its less than 10 dollars /year in my taxes. Specific missions like this one, around 1 dollar from that 10

Reply to  Warren
March 10, 2020 8:06 am

I hear that often about space missions, and fact is the spending for all of NASA is in the range of the margin-of-error of many other US goobermint expenditures. Hillary Clinton’s State Dept’s unaccountable “loss” of 6 billion when she was Sec of State was dismissed as a bookkeeping error, and that’s just one example of dozens. Rumor is now that missing 6 billion was actually part of the money laundered for Ukraine “operations”.

Reply to  Warren
March 10, 2020 10:21 am


IRS forms need to divide expenditures into 100 boxes. Minimal constant expenses such as the cost of the courts, congress and the Presidency, border protection, National Debt,etc, you know those things actually authorized by the constitution, are to be funded. For all other items, when you submit your tax forms, you determine WHERE you want YOUR money to go. I would, at the beginning , limit the maximum % at any box to 5% requiring any individual to fund 20 of the 100. of course things like the military would have multiple boxes due to the % of current spending, so in the end you could end up spending almost all of YOUR money on the military and NASA. As programs get defunded, the groupings would change, be more specific, and the size of government would shrink, because those actually paying for everything would be making the decisions.

I guarantee space exploration will get more money than they get now if actual TAXPAYERS decided where he money goes. If I could make that decision most of government would be defunded, but I am not a decider now, I live in Nevada where both senators are Democrats so my vote means nothing almost nothing because their priorities are essentially the opposite of mine.

You are just upset that this part of OTHER PEOPLES MONEY is not forcibly directed to where YOU think it should go.

The above system would soon see the elimination of the majority of programs created by politicians to get themselves re-elected. You know, the ones YOU probably want funded. The taxpaying American public know the value of space exploration.

As the original Star Treck began with: “Space, the final frontier…” What about that short phrase so you not understand?

Paul Penrose
Reply to  Warren
March 10, 2020 10:24 am

Not really. What they might be shocked at is how much has been given in tax rebates to rich people buying electric sports cars. Or subsidizing windmills and solar panels that produce very little usable energy, but contribute to destabilizing the electric grid and raising electric rates.

Right now humanity has all it’s eggs in one basket: Earth. We are only one big space rock away from obliteration. The only way to prevent that is to become a multiplantary species, and that can’t happen without a lot of research and investment in space technology and engineering. And since these technologies, once developed, benefit us right now, here on this planet (think GPS and weather observation for starters), it’s a win-win no matter how you look at it.

Dodgy Geezer
March 10, 2020 3:29 am

Their problem is similar to one I saw many years ago during an expedition to examine flora and fauna in the rainforest canopy.

The tops of rainforest trees are very high, delicate and similarly ‘lumpy’ and misshapen. The team were using a balloon to drift over the canopy, and an anchor to hold them in position over items of interest. They then let down a large but light framework – fibreglass rods and nylon mesh – which plated itself over the irregular surface, adapting to the bumps and bridging the gaps. The framework had hard points on it where the scientists could walk and examine items close up.

A similar structure could make contact with the asteroid, and hold the collecting gear safely about 20 ft up, from where observations and samples could be gathered….

alastair gray
March 10, 2020 3:45 am

Why dont Nasa get Gavin Schmidt to model it for them. After all we have entrusted the economic future of the world to the integrity and acumen of Gavin’s models so they just could not go wrong with Gav at the helm. At a pinch he could adjust the asteroid surface to the smoothness of a builliard ball and the the landing problem would go away.

Reply to  alastair gray
March 10, 2020 5:31 am

Gav will be useful for hindcasting where the probe actually was for the last 100y. +/-0.01mm . Actual tracking data will be “corrected” for biases until it fits his modelled trajectory.

Reply to  alastair gray
March 10, 2020 5:34 am

Oh wait , there’s not CO2. All Gav’s modelling experience is based on the fact the CO2 controls everything so probably would not be applicable to an asteroid with no CO2.


Jean Parisot
March 10, 2020 4:31 am

Does the software get a cool shirt to wear?

Paul Johnson
March 10, 2020 6:27 am

Shouldn’t the landing area be designated Terrain-OSIRIS-REx? (with apologies to Jurassic Park)

Reply to  Paul Johnson
March 10, 2020 7:12 am

While in college I was a small part of a team that was developing an updated monitoring system for nuclear power plants. The system was called OEDIPUS, so obviously the control panel was referred to as OEDIPUS REX.

Joel O’Bryan
March 10, 2020 7:40 am

With robotic sample and return missions becoming a reality there is no reason to waste vast sums on manned missions beyond LEO.

Don K
Reply to  Joel O’Bryan
March 10, 2020 9:05 am

I agree with you Joel. But it’s you and me against the rest of humanity who have seen too many Star Trek episodes and don’t understand just how difficult and costly it is to keep humans alive in non-terrestrial environments. And just how poorly designed the human body is for space exploration.

I can see a sort of vague argument for eventually putting a modest base on one of Mars moons to control robotic surface activities there without a 6 to 40 minute command-to-response latency do to the speed of light. But that’ll cost many trillions of dollars. And we’ll probably kill a few people doing it.

As for the moon. We can control robots there 24/7 from an Earth base and a couple of relay satellites. Even though we’ve had the technology to remotely explore the moon for half a century, we’ve done next to nothing. A couple of Russian efforts in the 1960s-1970s. And some recent Chinese and Indian efforts of which I believe only one vehicle (Chinese Yutu-2) is currently active

Smart Rock
March 10, 2020 9:21 am

This is interplanetary exploration; the kind of thing that NASA was intended to be doing, and I wish them good luck with it.

The rest of the world is heavily indebted to the USA and its taxpayers for all the space exploration, plus all the satellite data we get for free. Those of us who bother to think about where it comes from are grateful.

Jan E Christoffersen
March 10, 2020 9:22 am

How big is Bennu? No info on the NASA links.

Bryan A
Reply to  Jan E Christoffersen
March 10, 2020 12:35 pm

282.37 × 268.05 × 249.25 ± 0.06 m

Mean radius
245.03±0.08 m

Equatorial radius
282.37±0.06 m

Polar radius
249.25±0.06 m

Surface area
0.782±0.004 km2

0.0615±0.0001 km3

(7.329±0.009)×1010 kg

Mean density
1.190±0.013 g/cm3

Equatorial surface gravity
6 micro-g[4]

Rotation period
4.296057±0.000002 h

And a cool rotational GIF
comment image

Something WIKI is still useful for

Jan E Christoffersen
Reply to  Bryan A
March 10, 2020 4:46 pm


Thanks. The mean density is surprisingly low.

James Schrumpf
March 10, 2020 12:10 pm

The NFT sounds like an improved version of the Pershing II and early Tomahawk cruise missile guidance systems, that used radar maps created from Digital Terrain Elevation Data (DTED). I used to collect that manually, at the 30-and-100-meter resolutions, for the old Defense Mapping Agency back in the ’80s. Cool to hear the concept is being used for scientific research now.

%d bloggers like this: