Real-Time Physics Simulation Forum

From the Meqon web site:

"Continuous Collision Detection
The new Continuous Collision Detection avoids tunnelling for fast moving objects and resolves penetration with ease. Instead of implementing a rather expensive geometry sweep or 4D algorithms we have a unique solution that produces the same results or better in shorter time."

Does anyone know what this actually does?

Does anyone know what this actually does?

I am afraid I can't give any details here. Our method gives almost no performance overhead but it is not 100% safe. It is designed to handle all regular cases for both static/dynamic and dynamic/dynamic but not extreme cases involving extreme amounts of rotation, etc.

Dennis

Well it looks like continue collision will be a very standard feature on all physics SDKs

"Instead of implementing a rather expensive geometry sweep or 4D algorithms we have a unique solution that produces the same results or better in shorter time."

... and what would be considered better results than collision response at the first time of impact?

dcoming wrote: ... and what would be considered better results than collision response at the first time of impact?

The main unknown of their approach is how they 'estimate' the first time of impact, rathern then what to do with this time.

I guess they either subdivide the timestep until there are no large 'gaps' in the motion or they use an iterative approach like the pure linear version of raycasting against the Minkowski sum (Gino van den Bergen) or the angular extension (described in my draft).

Meqon has GJK, which makes a Parameterized Sweep less likely (some of the shapes are implicit without features to Parameterize), unless they add an additional explicit polygonal representation 'inside' the object.

Dan, what method do you use to estimate or calculate the first time of impact? And which shapes do you support?

I forgot to log-in. That last guest post was from me in case it was not clear.

Anonymous wrote:Well it looks like continue collision will be a very standard feature on all physics SDKs

Sounds great to me! (especially for some better human-computer interaction)