Hello folks! I’m casting a line out for help, since I’m in a bit of a tight situation. Me and another classmate in University have to make an addition to an existing AI class that we’ve been learning about in class, but there’s only roughly 11 days left in the project before it’s due. As such, I’d like to ask this forum for potential ideas as to what I (and my partner) could implement within a short period of time.The assignment allows the students to pick an area of focus and an ‘approach’ to the assignment, but because me and my partner are not particularly strong programmers and are quite new to Unreal Engine, I’ve narrowed it down primarily to an “Educational” approach, where the goal is to teach a simple concept to someone new to Unreal, as the process can also allow us to learn from it too.
The main form that the AI takes place in is an AIManager utilizing an A*Algorithm pathfinding system through a stage mapped out with BSP, done mostly in C++ with some additions in Blueprints. Ultimately, the point of the course is to emphasise C++ learning, so preferably I’d like to minimise using Blueprints in the solution. Currently, in-class the state of the AI allows it to Patrol, Evade and Engage a player by moving through nodes placed in the level, but it is quite a simple implementation - the enemy AI cannot turn to face the player character yet, and there’s some mild bugginess occurring in the AI’s behaviour regardless.
The following headings are the areas I believe that are most approachable.
“This can take any form that you wish, as long as the focus of the development is regarding the control (movement, sensing, decision making, communication etc.) of a non-player character”
-
Movement:
o Expanding on the A* pathfinding navigation to allow for ledge jumping
o Adding a cover system for NPC’s to hide behind obstacles
o Calculating flanking routes on the player. -
Sensing:
o Giving NPC’s well simulated sight, hearing, or smell
o Giving NPC’s a memory system for the player’s past location / actions (see popular stealth games) -
Decision Making:
o More behaviours in the Finite State Machine (FSM)
o Replacing the finite state machine with another Behaviour Selection method such as a Behaviour Tree or Hierarchical Task Network (HTN)
** I AM NOT ASKING FOR SOMEONE ELSE TO DO THE WORK FOR ME(Out of both academic fairness, and for the sake of my own self-improvement). **
TLDR; I’m asking for some suggestions from this forum as to what’s a simple or relatively straightforward implementation that me and a schoolmate can break down for educational purposes, that is doable within a week’s worth of work (taking into account multiple other classes and out of school work.) If links to video series that break down the implementation exist, I would like to see those as well so I can make my own version of those implementations.