Workbench Status Page
So how does my workbench status page work?
It’s a pretty basic setup, I used Tensorflow to create a model from a set of training data I captured with a small webcam mounted above my workbench (I use the term “Mounted” loosely as it’s mostly just jammed between the whiteboard and the wall).
You can see the camera sitting on top of the whiteboard, running into a Coral Dev board, which is running the Tensorflow model:
Again, take note my extremely high tech and professional mounting system of two thumbtacks holding it in place.
The coral board serves the API endpoint, which runs the current state of my workbench through Tensorflow and makes a drunken guess as to wether my workbench is clean or messy!
Is it totally live?
It’s close, it only checks once every 5 minutes. The crap on my workbench doesn’t really move any faster than that so no reason to check every single request.
Getting the image from the camera is ultimately the slowest part - which is ultimately why I just cache the results for 5 minutes.
Why don’t I include a picture of my workbench?
I work on some proprietary things and hardware that I have signed an NDA for. I really don’t want to have to remember to shut off the thing posting images or not, so I just decided not to include an image.
You just have to trust the drunk computer trying to figure it out on its own.
So why did I makes this?
I like to have a clean workbench and desk at the end of a day, just kind of makes starting the next day a bit nicer, so I kind of wanted to create a small thing to shame me publicly if my workbench was messy!
Also, it was just a fun little project. Why else do we engineers do what we do.
You can see it here but be warned, it’s a fun hack project so don’t expect anything amazing.
#machine learning #toys