Jumping wild circus mouse
This experiment started with setting things up to eventually try my conveyor belt maze. But I figured I'd first test my apparatus with a simple setup.
I have a "Raspberry Pi" model 3B+ credit card sized computer on the shelf, attached to a camera module with a long cable. The camera continuously records video clips and my software analyzes the clips to detect motion. Anything without motion is discarded. Previous experiments just captured still frames at 5 frames per second, the video feature is something I added in early 2019
The first creature that showed up was a shrew. Initially I thought I might end up with another mouse vs. shrew themed video. But the shrew only showed up a few times and never did anything interesting.
A little mouse showed up. It also didn't try too hard to get at the food. But the problem was, I started this at the beginning of October 2019, when outside food for rodents is still relatively abundant.
From my mouse trap experiment I knew mice could jump up to 30 cm high, but this mouse wasn't prepared to jump higher than 20 cm. Maybe escaping from a bucket trap was a stronger motivator than just getting at food.
So I experimented with more interesting jumping off point. The mouse was happy to jump off a stack of yogurt cups, but as the stack partially collapsed, it wasn't successful at getting at the food anymore.
But this is a very wobbly platform, which makes it difficult to hit the target. The platform moved too much to the side during the takeoff jump, messing up the mouse's aim.
Though the shrew totally owned my hole experiment and big maze experiment, if it had stuck around, it would have been totally out of its league here. With short legs and very poor eyesight, this sort of jump would not have been possible for the shrew. I'm quite amazed at what the mouse was able to do.
But the shrew returned much later (at right) but it only sniffed at things. I don't think shrews jump or climb much.
I should add, the mouse was able to aim for the target surprisingly well, but I kept more clips of it missing the target.
I had the camera running continuously from the beginning of October 2019 to mid January 2020, but my software discarded over 99.8% of the clips on the spot. I did have a few gaps in recording from revising the software mid-way and introducing bugs. These experiments always involve some amount of programming.
The mouse typically showed up for about 20 minutes a night, but some nights not at all, and sometimes during the day.
I kept modifying the experiment, trying to make things more interesting. Here's a food platform on a metal rod pedestal, on a base that is round on the bottom to make it wobbly. I hoped the mouse couldn't climb the rod, but it had no problem getting up. So this pedestal ended up mostly serving as a wobbly jumping-off platform.
I created a teeter-totter platform, with screws in the bottom to weigh it. It's on an odd shaped metal rod stand, with a big washer on the rod to try to keep the mouse from climbing up the rod to get to it.
The mouse was surprisingly good at jumping off the wobbly teeter totter.
The mouse was also willing to jump a considerable distance horizontally off the tall wobbly platform (tipping away from the recoil). The tall wobbly platform was a relatively better jumping off point due to its mass from the metal rod.
It was a fun experiment to run, every morning checking if the mouse
had visited see if it tried to do anything interesting. Mice normally
live very secretive lives, and to see what it had been up to in full
color high definition!
But my setup didn't capture sound. And even if it did, it wouldn't have been that interesting. So after spending much time selecting which clips to use came the long work of adding sounds for everything. I had a mess of plastic containers, scraps of wood, and various kitchen utensils on my desk that I used to make mostly believable sound effects to go along with the action.
This effort is justifiable if the resulting video gets millions
of views, but whether it will is always a gamble. Some of my past
mouse videos did much better than I hoped fir, but past success is no
guarantee of future success. Views of my old mouse videos
have declined quite a lot since early 2019, so I have doubts
but I figured I'd give it another try. You never know.
More mouse experiments:
Mouse experiment setup vlog (2018)
Mouse in a lego maze (2019)
Mouse trap maze experiments (2016)
Rat trap fail (2016)
To my Woodworking website.