This is my first blog post for my individual project I am taking in my third year at university. For the module I have until May to create a high level game with documents such as a proposal and report.
For this module I have chosen to create a real-time strategy game using Unity 3D and intended for a PC platform. The genre is similar to a couple games I have worked on in the past but I have chosen it to further explore AI in games. As RTS games have a lot of AI elements to them, ranging from some basic decision making to unit formations and AI pathfinding. As I have a strong interest in the AI, I want to make it has good as it can be, so I have decided to save some time in other areas for this. One example is making use of assets I have already acquired for a turn-based strategy game I made called Ethica Divided. These assets are medieval/fantasy genre and having them already saves time locating a whole new range of assets.
This blog post marks a rough start on the document side of the project as I start writing the MoSCoW Analysis section of my proposal. As this section helps me see clearly what areas are more important to work on over others to get the core mechanics of the game done before the added features. But a basic prototype was formed over the summer that is a strong basis for my project. It has a few core mechanics already implemented such as some basic building placement, unit movement and unit selection. The building placement and unit movement wouldn't of been possible without doing the AI pathfinding first, as this forms a basis to these mechanics, as well as many others in my project.
The AI pathfinding works by creating an A* algorithm. An area I know a lot about by using it in my turn-based strategy. It creates a grid of nodes (node being a world position and whether it is walkable) across a defined area so that units can find the shortest past from one node, to another. As a node also contains information about if that node is walkable it works well for building placement. Finally creating a pathfinding algorithm this way compared to say a baked NavMesh in Unity allows it to be dynamic and having nodes update to see whether that area is walkable again or not.
The next few weeks will be spent mainly on writing the proposal, but also researching into new areas to understand how to create some mechanics. Below is a screenshot of the prototype in it's current status:
This image shows 2 capsules representing units. One villager and one just general unit for testing. The top left shows you the currently selected unit and when the villager unit is seleceted, two UI buttons load up to build either a town centre or a house. When placing a building it will check to see it can be placed in that location, the two pink areas on the left are examples of non walkable areas where buildings cannot be placed.