I'm a small electronics engineer and I've actually put a reasonable amount of effort in to figuring out how they could fix this problem. It is possible to solve more or less as people are describing (lots of questions: where to put the readers, if it's in the posts what happens when they move, in the ice, passive/active tags, etc). The problem becomes getting more accuracy than cameras. To do this you would need sub mm accuracy. This isn't impossible, but it's not easy especially if your plan is to have a totally passive puck.
At the end of the day this project is probably something the nhl needs to fund a school/Ph.D. Student for as it probably isn't worth a ton of money on its own. If there's an application for this technology outside of hockey someone needs to develop it independently and sell it to others.
|