Bennu Mappers – FAQ

Yes! We are providing nightly summaries of everyone’s efforts to the folks working on OSIRIS-REx at the University of Arizona, and they are using your data to look for the safest places to possibly grab soil/pebbles from Bennu.

You aren’t the only person who will mark an image. Every image is given to at least 15 different people. Your combined data is what we use. We assume that mistakes are all going to be different, and correct answers will all be similar. With Bennu, the variations in your data and the other data of other people also helps us understand more about the asteroid. If everyone marks the same long axis of a rock, it tells us the rock is really elongated. If everyone marks a different long axis, and the lines mostly cross in the same place, it tells us the rock is pretty round. By having a bunch of different people all look at every image we get better data than we could get if only one person looked at an image, even if that one person was amazing!

The software did save your data! When we first started seeing this problem, we asked people to tell us their usernames so we could look in the database. In every case, the data was correctly saved! For reasons we haven’t figured out yet, the software just sometimes doesn’t give you a new image correctly. You can safely refresh the page to get a new image, and know that everything you did is stored for the OSIRIS-REx team to look at.

Itokowa as seen by JAXA’s Hayabusa spacecraft. It is much smoother than Bennu!

When we wrote this software, we didn’t have images of Bennu yet. At the time, we thought Bennu would look a lot like the asteroid Itokowa. Unfortunately, Bennu had other ideas! Bennu is much rockier and its craters are much harder to see. For now, we aren’t going to update the tutorial (because we want you to be able to do it quickly!) We are going to post more examples for you to look at to understand what to look for.

There is no hard line scientifically. The rule of thumb (or rather of blue dot!) that we’re using is as follows:
smaller than 1/2 blue dot, too tiny to mark
1/2 blue dot to 3x blue dot size, rock
3x blue dot size, boulder

The rocks have self-identified and want to be called their chosen names.

No. We won’t let you have a zoom function because we like you. If you had a zoom function, you would try to mark dust, and get super fussy with how you mark things, and might never finish an image. We want you to succeed and feel good about what you’re doing, so you don’t get to zoom. You are seeing things the same way the science team looks at things when they mark images. If they think it is good enough for them, then it is good enough for all of us!

Not going to lie; mapping Bennu is tedious. That said, you need to mark most things, but not everything! Glints of light just a pixel in size? Don’t mark. Textured dirt that might be rocks, but is impossible to tell? Don’t mark. In general, mark everything larger than 1/2 a blue dot. It takes me (Pamela / startryder) anywhere from 5 minutes to 30 minutes to mark an image. The images with HUGE boulders go way faster. The ones with hundreds of small rocks… This is tedious work. Which brings us to the next FAQ question…

Sometimes doing science is really tedious. Mapping Bennu is that kind of science. Can we recommend listening audiobooks and podcasts? Need specific recommendations? Ask over on Discord.

We have our own 365 Days of Astronomy Podcast which brings together a bunch of cool astronomy content. Check out the link in the sidebar.

If software could map Bennu, we would totally have software do it. Unfortunately, software that is used for visual tasks (like face recognition and rock recognition) requires training data. We didn’t know what Bennu would like until we got our images (delivered May 15, 2019!) This means there is no training data. Someday, all the mapping you are doing may allow us to train an AI, but for now, you are our only hope. Your eyes and clicks are the best current way we have to map all our images quickly.

After July 10, the OSIRIS-REx mission will select areas for more detailed images. They are going to use your results to select these areas! This will let them go from a bunch of possible places to get a rock sample, so one place they will get a rock sample.