Thanks for the additional explanation. This makes a lot more sense to me. I can see how this can make a big difference in our understanding of real-life, real-time, real-location usage. As has been suggested, this collected, anonymous info can then be used for faster “dial-ins” or having providers getting closer to anticipated usage the first time.
I’m still curious as to how this will play out. Please keep us posted.
Thanks!
You’re ignoring this member’s posts.