Campus Navigation Game

My dissertation project as part of my BSc (Hons) Computer Science degree from Lancaster University.

Grade: A

An open-world videogame set on the Lancaster University campus, designed to help incoming students learn the campus layout before arrival.

Read my Dissertation!

Aims

• Create a videogame modelled on the Lancaster University campus
• Videogame should encourage repeated exploration of the virtual campus
• Leading to repeated exposure to campus landmarks
• Conduct a user study on the impact of the videogame on navigational ability

Introduction

• Existing campus orientation techniques are flawed
• In person orientations in Welcome Week lead to information overload

• Students often rely on GPS systems to navigate

• Explored virtual campus tours as an alternative

• These do not have any replay incentive

• Videogames were the solution

• Use implicit learning

  (knowledge) acquired largely independently of the subject’s awareness of either the process of acquisition or knowledge base ultimately acquired

Problem Statement

• The campus navigation game shall contain a 3D map modelled on Lancaster University campus.
• The map shall contain recognisable landmarks from the Lancaster University campus.
• The campus navigation game shall allow the player to control a character that can walk around the map.
• The campus navigation game shall include mechanisms that encourage the player to explore as much of the map as possible.
• The campus navigation game shall not focus on the map being a campus.

Game Design

Much research and consideration went into each game design choice. A brief overview of some of these choices are in the following section

Game Genre

• Open world vs Linear
• Chose open world, as this supports spatial memory of map
• BoTW and Stardew Valley researched as inspirations
• Farm life simulation loop of Stardew Valley provides structured exploration of the map
• Quests can also do this

Perspective

• 2D/HD2D does not match real perspective of student on campus
• 3D best option

Art Style

• Pixel art not suited to 3D
• Low poly style chosen for:
  • Strong performance
  • Abundance of free models
  • Semi realistic, meaning game will reflect reality of campus

Game Loop

• Aim to make game replayable, to ensure repeated exploration of campus

• Initial ideas of a student game loop, with dorm customisation

  • But I wanted the campus not to be the focus - learning should be implicit

• Cooking game with delivery service

Landmarks

• Due to time constraints, the campus could not be modelled to perfection
• Subset of key landmarks should be modelled in more detail
• These were chosen on relevance and recognisability
• Used Blender

Game Engine

• Choice between dedicated game engine vs a web framework
• Game engine provides ease of development and more complexity, compared to the performance of a web framework
• Initial testing done in Unity, but had to move to Godot due to licensing issues

Cooking Minigame

• Need for replayability

• Customisation of restaurant came up as a possible idea

  • This takes too much time away from exploration

• Delivery system + short minigame settled on as best option

• Explored existing cooking games such as Overcooked and Papa's Pizzeria

• Simple but addictive

• Fruit Ninja another inspiration
  • Fast paced, leaving more time for exploration
  • Clear visual feedback
  • Combos

Delivery Quest

• Inspiration from Wii Party
  • There is no direct use of GPS, forcing player to learn layout

• Provides structured navigation

Replay Incentives

• Tools
• Recipes
• Vehicles

Quests

• NPCs make game feel full, like the real campus
• Quests from NPCs to unlock tools, recipes, vehicles
• Provide semi-structured exploration of the campus

Inventory and Crafting

• Inspiration from Stardew Valley and Minecraft
• Provides more encouragement to explore map, as they have to find items to craft with

Testing

Testing showed satisfactory performance with some stutter

Evaluation

Evaluation Requirements

• Evaluate landmark recognition
• Evaluate navigational ability

Evaluation Plan

• Users had to identify landmarks along a route, like in this picture

• They did multiple of these tasks before and after playing videogame
• See if videogame improves navigation

Evaluation Metrics

• Measure precision, recall, F1
• Measure time taken
• Measure user experience
• Narration to see how navigation changed

Results

• There was a slight increase in the metrics, suggesting landmark recognition and navigational ability improved
• However this could be the learning effect
• There was also a very small sample size of 2.

Improvements

• System Usability Score (SUS) was only 50

• Most complaints were about the tutorial lacking

• More focus should have been on realistic campus modelling, rather than game features

  • My lack of Blender experience was a detracting factor

• Cooking minigame distracted from navigational tasks

  • Should have focused solely on parts of the game that force users to explore map

Overall Takeaways

• Game partially met objectives
• Showed that the concept of a videogame for campus orientation is possible
• Could be developed further, with a more realistic map and polished mechanics
• Larger scale study needed to provide concrete results
• Game should be simplified and focused on navigation aspects