Many Thanks for placing such a unique bid. I have lots many things to write after reading and doing some research to understand what we are trying to come up with.
First I would like to share that I am new to freelance but not to development. I am delivering solution since last 15+ years and since 2012 working on big data and analytics projects. I have delivered solutions using Hadoop, R and Python on large data sets (upto 6TB). Worked with Text Analytics (Batch and Real Mode), Image Analytics, Data Mining, Data Scraping, etc. I have implemented many machine learning algorithm like Random Forest, kNN, Neural Network, Euclidean, Pearson, etc. Carrying my learning I would like to be part of the development of this system and continuing would like to know more. I understand the reason of opting Watson (an impressive API framework and you are partners to IBM) but IBM also offers Video Analytics separately. Have we evaluated that? Also knowing we have OpenCV like open source available that can easily resolve the purpose of identification with real time video streaming, what challenges do we think we may encountered if we opt this?
I am extremely impressed with the noble thought behind this tool and also gone through the attached PDF. I read about prediction and would also want to know more on this? Considering "ELSAG Automatic License Plate Recognition" sounds like solution is supposed to be build for North America....
Please share if I am allowed to discuss this further.