upgrading trash skimming robot

[pearlxoxo]

hello! so i made it to regionals with my prototype of an rov with a net attached to collect trash in waters, but the teachers that judged my project worried about wild life getting caught in the net. my teacher wants me to upgrade my prototype with this feedback and i could use some suggestions. one judge told me to use a rotator to detect what is solid and what is wild life but im a bit confused? i went looking for rotators and they didn’t look like they matched what he suggested.

[amyCC]

Hi - Congratulations on advancing to the regional fair! That's wonderful.

I will let other experts reply with suggestions for the question you asked, but I wanted to say congratulations.

Amy
Science Buddies

[MadelineB]

Congratulations on making it to the regionals! Like Amy, I want to say congratulations and will leave answering your questions about detecting wildlife to other experts.

I caution you to double-check the specific rules for the regional fair ... make sure whether they allow you to modify your project! I know some regionals have strict rules that the project you show at the regional should not be modified.

However, in either case, you could certainly be prepared to talk about proposed changes ... that you might apply to next year's project!

Good luck!
Madeline

[bfinio]

Hi -

I'm also not really clear what a judge meant by "rotator." I can think of a few different ways to potentially detect wildlife in the net, but they would be significantly more advanced than the baseline ROV project that we have.

1. Use a camera with computer vision. This would require another board, like a Raspberry Pi, because the Arduino is not really powerful enough to do computer vision on its own. If you browse the artificial intelligence area of our website: https://www.sciencebuddies.org/science- ... telligence you will see multiple projects about computer vision. So, for example, you could train a model to tell the difference between pictures of trash and pictures of fish. But again, doing that on a regular computer is one thing (that's what our project instructions tell you how to do), doing it on a mobile robot is more complicated.

2. Don't use a camera, but try using other sensors to detect the difference between inanimate objects and wildlife. For example, you could mount an accelerometer on your robot (we have a tutorial here: https://youtu.be/zpV7ac3ecl4?si=9NMLGIdTRkeptjag) and periodically stop the motors so the ROV stops moving. If the robot is just holding trash, the accelerometer readings should be pretty low (assuming the water isn't too choppy). If there's an animal caught in the net, flopping around and trying to escape, the accelerometer readings would be higher.

3. You could add safety features to try and scare animals away from the robot to begin with. The motors are already somewhat loud, but you could add buzzers or speakers to make it louder, or maybe flashling lights (although I don't know if lights would attract some animals like fish).

Hope that helps! But before doing any of that, do make sure you check whether you're allowed to modify the project, like Madeline mentioned.

[pearlxoxo]

Thank you everyone for the congratulations! I checked the ISEF rules for the regional science fair and it basically says that after I compete in an affiliated fair then projects may not be changed or amended. I’m going to double check with my teacher to see if what qualified me for regionals is the affiliated fair and then ask to see if changes are allowed. If not, that’s cool. And I will keep in mind the options that bfinio gave me . Thank you all for the help!