Predictive Algorithms Instead of Reactive Solutions

[jast]

Continuing the discussion from PCDS Continued (Persistent Cat Deterrent System):

Hope @Clayton won’t mind me spinning off another topic so as not to harsh the mellow on the thread from which this is spawned.
Jim asked about the speed of the IP camera’s processing for detection of motion, and whether that would be fast enough for the intended deterrent to be effectively deployed. Although the assertion is that this happens “in real time” I still question the ability of ANY reactive method for deterring a critter who has found the fun in “running by”. The thought struck me: if you’re moving to an IP camera and processing power anyway, could you, instead of being reactive to detected motion, use the processing to monitor the motion outside of the intended fur-free zone, process that predictively, and, with some accuracy, deploy deterrent BEFORE said critter is actually “in the moment”?
Assuming we are working with this scenario, still

Help with co-habitation using lasers

Hi Electronics! I am in a bit of a conundrum. My three best friends (Wife, Cat1, Cat2) co-habitat…but barely. Would anyone be interested in helping with a project? When we are done, it could be turned into a class to earn honorariums for the committee. I’m fairly handy, can do some basic coding, and own a couple of arduinos. I even understand some electronics basics. For Christmas, I got my wife a few new pieces of furniture… My wife thinks furniture is for sitting, relaxing and aestheti…

where we have a choke point (of a fashion) we could monitor the traffic OUTSIDE the choke and, I wonder if there already exists or could be relatively easily created, an algorithm which might reliably predict whether Cat1 or Cat2 is headed into the compound (as opposed to a casual drive-by recon mission). Another upside (I’m guessing) is that the same algorithm could sense Spouse1 and/or Spouse2 and/or, should the need arise, Childn), probably based on size of signature alone, and choose to not deploy said deterrent (or whatever we choose to do with the information this hypoteical alogorithm provides).

Anyone have thoughts on that?
Meanderings welcome. This is a muse thread.

[Clayton]

@jast Thank you.

With the IP camera, it becomes more preventative…instead of waiting for the cat to cross the threshold, I can designate zones…so even if the cat crosses an imaginary line 30 ft away, I can trigger the deterrent, therefore eliminating his ability to jump past the deterrent before it can react.