Quote from: M @ Gluon on January 31, 2018, 06:22:58 PM

Quote from: willoweb on January 31, 2018, 05:29:54 PM

The implementation of the idea of your project feel particularly strong in business with large fleet of vehicles. Because as far as I understand the system when connected to the vehicle and reading and transferring data to the cloud will allow, after analysis of historical data of similar vehicles to predict the damage indirectly.

You are absolutely right, and there is so much more.
The assistance and financial savings any size fleet can enjoy with our system is significant. Just as you say, as our AI develops knowledge from statistics of the users particular fleet and vehicles, the ability to predict and prevent damage will get better and better with time, the right spare parts can always be at hand (and ordered trough our system) and vehicle down time will be minimal. Or if any vehicle is down for unforeseeable reasons, alert and rescheduling to cover for that vehicles suspended service will be seamless at the administrations fingertips, and so on and on..

You will use your own equipment or rent machine time for training neural networks? datasets will be collected from users and depersonalize for training neural networks? perhaps someone will want to share telemetry data of his car. This is certainly extreme, but such people can exist. on the other hand, after 5 years of collecting and studying such data and AI will be very good.