Free Artificial Intelligence Path-Finding Methodology Essay Sample
Computer games came up on the scene as early as 1950s. With time the game developers revolutionized the means of game production to make them more interactive, more realistic and more captivating for the audience. These games were made for an audience of wide range of age, IQ and interests. The study of AI and its adaptation in programing of modern games and simulations gave a more human feel to the overall gaming topology. This was possible as AI games were not strictly based on logical iteration, had it been so then while playing ‘First Person Shooter’ type of games; the human would not simply stand a chance against the Non Player Character (NPC) i.e. the computer. With time more strategic games like Command and Conquer, Red Alert and Generals came up, they exhibited a wide spectrum of AI utility in multiple aspects as well as methodologies.
The early employment of AI in games emerged from the concept of ‘scripting’. This was defined as multiple set of options for the NPC to exhibit control and behavior. With time new techniques came up, such as the Pathfinder. This technique is more employed in strategic game development in which an onscreen NPC is to navigate its way from one point to another on the game’s map zone. The development of the Path finding technique takes into consideration multiple aspects like location and behavior of different objects in the game zone e.g. living or dead obstacles, type of terrain, explored areas etc.
The Path finder technique with time grew up into a subfield of AI termed as ‘Navigation’. A simple path finding technique is focused to guide our NPC entity form one point to another, however, when a lot of information or map layering is enforced, now to find out the best possible way as per the game criteria to score a kill or any point of interest, the NPC has to be told what all is there around him, or between him and his target. Feeding of said info to the NPC is possible owing to the AI’s Navigation technique. Further to the same, as the game progresses the schematics and dynamics of the game’s map also keep changing e.g. if at the start of the game there was a building between two entities, the same building may get destroyed with time and now the NPC can walk through over the debris of the building. Any such changing situation leads to the requirement of Dynamic Navigation in the gaming environment.
Path Finding Methods and Dynamic Navigation Mesh Generation Techniques
In any virtual world like an AI game or a simulation the first problem faced by the devleopers is to tell there entity or NPC about the world around it.(Hale, 2009) The NPC in a game is not blessed with the five senses like humans, rather it has a more restricted approach, he has to be told about everything, using said topology all the objects recognizable along with their attributive properties are taken into account as the positive objects and there location or the area of effect is taken as the positive area. What remains on the map interms of area either in 2D or 3D is taken as the negative area. The prime obejct for the Path finding technique is to navigate the object through said negative area. However the NPC has to choose something known in the gaming world as a ‘Good Path’, this path is selected keeping the physical limitations of the NPC in mind e.g. its fuel, time, type of movement mechanism,money and distance. Moreover the physical properties of the negative areas of the map are also to be known that may have varying effect with respect to the limitations of the NPC. As a solution to said problem different PathFinder and movement algorithms are developed. (Amit n.d). However with time it was noted that a path finder approach if based on a point to point system would render the entity cludeless of what lies on the map apart from the points that are described to it. Therefore another system of Navigation based on a series of polygons was adopted (Navigation Mesh Reference n.d). In said path the NPC is directed to walk a path that is represented by an array of polygons, and the NPC can walk through said polygons to reach its goal, this in a way tells the NPC what all space on a map is walkable. This enables the NPC in a game to accompalish free movement around the corners, and taking sharp turns etc. There are multiple techniques for said mesh generation, that would feed the entity regarding the walk paths, the obstacles and other states of interest in a gaming environment.
AI middleware packages available to assist with navigation mesh generation and AI systems
As the art of computer gaming has revolutionized recently, there are number of prefabricated software packages already available of the shelf that may be employed by a novice game developer to accomplish the simulation of different behaviors by its NPC. Few such middleware available in the market are described below.
Kynapse is a middleware software product by Kynogon. It includes an automatic AI data generation tool, a complete 3D pathfinder, spatial reasoning, management of dynamic and destructive terrains, streaming mechanisms to handle very large terrains, optimizations for multicore, multiprocessing and cell architectures. Said middleware has been utilized for development of AI games as well as real time simulation systems. (Kynapse, 2011)
According to Alex, Havok game developers have come up with a new AI middleware library. (J.Champandard, 2009) The prime features of interest comprised of a very fast core navigation mesh building algorithm, real time dynamic change adaptation, “fully extensible and customizable path-finding solution, a hierarchical pathfinder that is multithreaded and platform optimized for all key gaming platforms” (J.Champandard, 2009).
Navpower is another high quality dynamic navigation tool that has been incorporated in many advanced games. It comes with an automatic Navmesh generation tool and simple API. Its build process generates a polygonal Navmesh over the complete map thus relaying the walk-able paths in a game level using level geometry as input. (NavPower Leading-edge AI motion planning n.d.)
A number of such middleware AI Navigation mesh generators are thus available, there applicability and compatibility with respect to the gaming platform, amount of memory, speed and accuracy of path finding laydown the rules for there choice.
Artificial Intelligence is a science for the present and the future. What supports our present day games and simulation software may prove to be the heart and soul of the next generation robots. Mankind is yet nearing the day when it might be able to inject senses like emotions, feelings, depression and happiness in its machines. If it is going to happen one day, it will surely be attributive to AI and dynamic navigation, however in a different perspective.