Project Title: Escape the Maze

This project involves the development of a competitive game where multiple AI agents, created by teams of students, navigate through programmatically generated mazes. The student teams are responsible for both developing the AI agents (clients) and setting up a server that connects these AI agents. Additionally, the server will load the programmatically generated maze maps and function as a viewer to display the progress of the game.

The goal is to determine the most efficient AI solution through direct competition among the AI agents, with the current positions of the players being highlighted on the screen. The player whose AI exits the maze first is declared the winner. The server will monitor win conditions and award points accordingly.

• Visibility: Visibility: By default, each player can see a 5×5 area around their current position, indicating possible directions of movement: North (N), East (E), South (S), and West (W)
• Turn-based Movement: The game operates on a turn-based system. In each turn, the AI will send a sequence of up to 10 steps (comprising the characters N, E, S, W) to the server, which will then move the player in the specified directions. For each command that cannot be executed (e.g., due to hitting a wall), the length of the allowable sequence for the next turn will decrease by one.
• X-RAY Points: At the start of the game, each player has 10 X-RAY points. These points can be used to access an expanded view around the player. For example, an X-RAY command, of 2 will consume 2 points and expand the visibility window from 5×5 to 7×7. The X-RAY command counts as one of the steps the player is taking, and are sent to the server as X, in the character array.
• The maze can contain additional tiles with various functionalities, which can only be placed alongside the path and not the wall:
  • Fog tile: The player can see an area of only 3×3 around him (X-RAY points can still be used)
  • Tower tile: The player can see in an area of 7×7 around this tile (X-RAY points can still be used)
  • Traps, which can be seen at a max of 2 steps away from them, and are of multiple kind, with a number n assigned to them:
    • movement decrease: decreases the maximum number of steps you can send to the server in the next turn by n
    • rewind: the players last n moves are undone
    • pushforward: the player is forced forward based on the direction of their movement for n steps
    • pushback: the player is forced backwards based on the direction of their movement for n steps
  • Portals: They work in pairs and connect one point of the maze to another.
    • Each pair has a specific id to represent a portal
  • Consumables:
    • X-RAY point increments

The maze must be generated according to the following constraints:

• The maze generator must function procedurally, creating mazes with a single entrance and a single exit.
• The entrance and exit of the maze can be placed anywhere in the maze, not necessarily on the borders.
• The minimum rectangular path from the entrance to the exit must cover at least 50% of the total rectangular area of the maze.
• Teams will choose the horizontal and vertical dimensions of each generated maze within specified maximum allowable ranges.
• The maze should be able to be generated in one of three ways:
  • completely random, based on a fixed seed, which may or may not be provided, and a set threshold that represents the maximum of special tiles that can be generated of each type.
  • semi-random, based on a fixed seed and a set number for each special tile that can be generated.
  • from an input file, which is represented by an image.
• The maze can never place two special tiles next to each other, or next to the entrance/exit

Each generated maze needs to be checked for validity to see if it respects the constraints imposed.

The maze generator must output an image, in 8bpp, grayscale format, with the following color representations for each pixel:

• 0 - wall
• 255 - path
• 64 - entrance
• 182 - exit
• 16 - X-RAY point increment
• 32 - fog tile
• 224 - tower tile
• 96-100 - trap movement tile with n = [1,5]
• 101-105 - trap rewind tile with n = [1,5]
• 106-110 - trap pushforward tile with n = [1,5]
• 111-115 - trap pushback tile with n = [1,5]
• 150-169 - portal ids, each value should only appear twice in a maze representing that portal pair (a maze should never have more than 20 portal pairs)

The maze generator can take the same image back as input to generate the exact same maze, or other images to generate new mazes. Obs: Images which contain undefined pixel colors are to be rejected.

Each request to the server is done by sending a JSON with the following format: {input: “string of commands up to length 10”} The server will output back a JSON with the following format:

"command1": {
  "name": "name of command, ex: "N"",
  "successful": "0|1",
  "view": "string of the matrix representation of the visible area around the agent after the move;"
          ex for 3x3: "[0, 255, 255; 0, 255, 0; 0, 255, 0]"
},
"command2": {
  "name": "",
  "successful": "",
  "view": ""
},
...
"command10": {
  "name":"",
  "successful": "",
  "view": ""
}

The server can store generated mazes as images and output them back on request.

The viewer should output the maze and the agents solving it in the following manner:

• 1920×1080 resolution
• 20 pixels/tile by default
• Mazes that don't fit on screen should have a scroll option.
• Zoom function
• Two possible viewing modes:
  • view the entire maze
  • view what each agent has explored so far
• The colors for each tile are left to the discretion of each team to make the output more interesting.
• Traps should have their value n on top of them if possible.
• The walked path should be represented by a solid line, and the planned moves of the agent by a dashed line.

An agent can work in one of two modes:

• real time: it sends the move commands to the server, receives back the success fail results and immediately follows with the next list of commands
• await for input: sends the list of commands, receives the results of the execution and awaits for user input before sending the next list of commands

Each agents performance is measured in one of three ways:

• Least time taken to solve the maze
• Least number of turns taken to solve the maze
• Least number of moves taken to solve the maze

For the real-time mode the agents will have a maximum time allotted before sending each command. If the allotted time expires, the agent is timed out and disqualified, and the maze is considered unsolved. The maximum time can be set before each run, or be preset depending on the maze difficulty.

After the initial phase of development the project For a successful collaboration, the teams should consider the following:

• collaborations will happens between teams in each lab
• meetings will happen between 2 members of each team
• each team can test each other's agents and offer feedback and suggestions

Alongside the normal team meetings, teams will meet between themselves with a limited number of persons to collaboration happens between teams in each lab meetings between 2 members of each team (potentially project manager/team lead + one other person) each team can test each other’s agents and offer feedback and improvements

The project is worth 6 points and is split into 3 major milestones, which will happen in Labs 3, 8, and 12. The milestones and their allocated points are as follows:

1. Setup milestone (0.5p), lab 3, 21-25 October 2024
   1. team members and their respective roles
   2. the chosen programming language, potential tools etc.
   3. the chosen development methodology (ex: Agile, Scrum etc.)
   4. The teams will have to create private projects on the MPS gitlab, in the form of Day_Hour_TeamName, ex: Monday_8_Team1
      1. All documents for the next milestones will be part of the repository for each team in wiki or others files.
      2. The README of the repo will contain all other links used - if the team uses a third-party solution for project tracking.
      3. The README will also contain the names of all participant members and the student groups they are a part of.
2. First demo solution (2.5p), lab 8, 25-29 November 2024, which will be graded on the following:
   1. Maze Generator (0.3p)
   2. Server (0.3p)
   3. Viewer (0.4p)
   4. AI Agents (at least one, points awarded based on the agents solving efficiency) (0.6p)
   5. Documentation, which must contain the following (0.6p):
      1. SDD
      2. testing reports
      3. meeting minutes
   6. Respecting the chosen methodology(0.3p):
      1. respecting team roles
      2. respecting scheduling and set tasks
3. Final demo solution (3p), lab 12, 8-14 January 2025, which will be graded on the following:
   1. Potential improvements and bug fixes on (0.5p):
      1. Maze Generator
      2. Server
      3. Viewer
   2. Improved AI Agents based on team collaborations (at least one additional solution different from the original developed one) (1p)
   3. Documentation (0.7p)
      1. new testing reports
      2. new meeting minutes (collaborative + individual)
   4. Presentation highlighting the results, potentially the evolution of the solution (0.5p)
   5. Respecting the chosen methodology (0.3p):
      1. respecting team roles
      2. respecting scheduling and set tasks