Network now...NEWS
Havok Adds A.I.
Intel-owned Havok, best known for its Havoc Physics engine, is now offering an artificial intelligence solution.
Havok A.I. promises unique solutions for pathfinding and character interaction in today's "highly dynamic" game environments.
The A.I. solution is available as part of Havok 6.5 toolset.
According to a Monday press release, Havok A.I.'s SDK features:
- A highly optimized and robust automatic nav mesh generator. Optimal nav meshes are generated in seconds from complex game levels of hundreds of thousands of triangles allowing for rapid iteration on level content.
- Fully integrated dynamic pathfinding. As opposed to being a static pathfinding solution with an additional layered dynamic avoidance technology, Havok AI is built from the ground up to work with dynamic environments. Its unique, innovative solution handles thousands of moving obstacles in real-time with high fidelity.
- Fully extensible and customizable pathfinding solution. The product includes a hierarchical pathfinder which is multithreaded and platform optimized for all of the key gaming platforms.
- Advanced predictive local steering module that complements the dynamic pathfinding capabilities. Characters predict the movement of obstacles and adapt accordingly, moving plausibly through the complex and often congested situations that arise when environments become dynamic.
- Out-of-the-box integration with Havok Physics, Havok Destruction, Havok Animation, and Havok Behavior.
Contact Us / About Us / Advertise With Us / BeltwayBlips / TechBlips / MassiveBlips / GamerBlips / FAQ
Subscribe To Edge Magazine / Corrections / Terms & Conditions / Privacy Policy
Copyright 2006 - 2009 Future Publishing Limited, Beauford Court 30 Monmouth Street Bath, UK BA1 2BW
England and Wales company registration number 2008885
Subscribe To Edge Magazine / Corrections / Terms & Conditions / Privacy Policy
Copyright 2006 - 2009 Future Publishing Limited, Beauford Court 30 Monmouth Street Bath, UK BA1 2BW
England and Wales company registration number 2008885
The features described mainly focus on navigation and pathfinding and little was mentioned about AI behavioral patterns and interaction with the player.
This is mostly a good thing. It means developers having to license less third party products to be up and running and perhaps saves some developing their own in-house AI systems altogether. It's success will be down to how flexible it is in serving different game genres and the breadth and range of NPC behaviors made possible (which will surely become broader as more devs tinker with it and find its limits).
Perhaps the only draw back might be this "standardization" of AI - it could cause many games to be even more similar to play through, especially if some devs are lazy and just employ basic out-of-the-box AI behaviours.
Well said. It's standardisation I'm worried about - just look at Unreal Engine.
There is nothing wrong with the Unreal Engine or the fact it's used by plenty of people. The issue is more that people look at the success of games like Gears of War and decide to emulate their look. That's what leads to standardisation.
One only needs look at the list of games using UE3 to see that some have managed to create worlds that look and feel far different to the likes of Gears: for example, take APB, Mirror's Edge, or even something like XBLA's Undertow.
Yes, some have. Some. You only have to look at the long, long list of below-average games which look nigh on identical to realise that maybe UE3 standardisation isn't all roses.
I'm not suggesting it is all roses, far from it; what I am saying, however, is that the poor workman shouldn't blame the tools.
My point is that a lot of these games would perhaps not have seen light of the day if it wasn't for UE3.
Sorry, but UE3 isn't the only middleware engine out there. To make such a claim is more than a little naive.
I never made such a claim. But personally I would prefer companies to use their own engines, rather than the likes of UE3 or the Cry lot. I think that would drive innovation more than relying on middle-ware solutions.
Stick to physics guys, please.
Any particular reason they should do that? They're clearly clever people at Havok and their physics solutions have allowed dev teams to stop reinventing the wheel and therefore reduced costs, and no doubt allowed people to concentrate on gameplay programming. Surely something that can help improve the general quality of AI (an oft weak aspect part of many games) is a good thing, no?
Mostly because I'm not sure they will deliver. AI is so integral and unique to a lot of games that it needs to be built ground up to suit said games. Having an out of the box system may hamper progress into AI with lazy devs choosing to settle for what Havok will provide instead. That's why.
Physics on the other hand, well... physics is physics, by definition it is a set of universal rules which games incorporate. Having a ready to use tool helps rather than hampers there.
I don't think you're quite seeing what Havok are looking to offer. This isn't a one-size-fits-all AI solution to replace all programming on projects, they're providing the framework and foundations for solid AI, focussing on path-finding through environments. Path-finding is perhaps the most common of AI routines and can be solved with a fairly generic solution, and then customised by the developers. To claim a solution like this will be to the detriment of games is the claim the same of all middleware, which is just not true.