MDMAniac said:
scottie said:
Position = acceleration * t^2 + initial velocity * time + initial position, where pretty much everything there is a vector.
|
The problem with this method is that it's computationally unstable, i.e. quite error-prone. Consider this, if accelerometers were any good for positioning, we'd have no need for GPS or GLONASS. Even for indoor positioning, accelerometers alone won't be enough.
|
Bit of a difference between
* Something that needs to stay accurate over 10 years, and the consequence of failure could be as serious as directing a person the wrong way down a one way street.
* Something that needs to stay accurate over 30s -2 hours, depending on the game, and the consequences could be as serious as making people have to compensate for the drift in their game of virtual table tennis.
I do see your point, but you are drawing unfair comparisons to prove it.