[p1esk]

I'm not very familiar with their theory and technology

It's strange that you're not familiar with an actively developing, widely known, 15 years old open source project with similar goals as your initiative. It's like if you wanted to create an OS similar to Linux, without bothering to learn much about Linux first. As an example of what I mean, you mention things like "brain is digital" and "combinatorial space", which seem to be rediscovery of Numenta's SDR. Also, I don't think your way of processing an object in different contexts is fundamentally different from Numenta's, based on what you described, however it's hard to say without looking at your code.

To avoid reinventing the bicycle, I suggest you engage in some discussion on Numenta's forums, or at least read their papers. It's quite possible that you might discover something they missed. But, if you don't know what they did, you will end up discovering a lot of what they haven't missed.

[0x7cfe]

When I said that I'm not familiar with Numenta meant myself only. My colleagues, that are much more experienced than me, do indeed know that project.