Group Projects

Group Projects

Overview

1. Arrive at farm
1. Arrive at farm
2. Barn Puzzle
2. Barn Puzzle
3. Follow Train tracks
3. Follow Train tracks
4. avoid enemies
4. avoid enemies
5. Chased by enemy
5. Chased by enemy
6. Factory Puzzle
6. Factory Puzzle
7. Traverse Forest
7. Traverse Forest
8. Cross Battlefield
8. Cross Battlefield
9. Arrive Home
9. Arrive Home

Puzzles

Barn Puzzle breakdown

I wanted the whole level to be a diegetic experience without UI or explicit onboarding. In order to support this, I kept interactions simple and readable, with minimal friction.

The player is limited to a interact-action that pulls levers or grabs boxes.

This barn was designed for immidiate clarity. When entering, the player can see both the blocked exit, and the new path above it in the same frame, establishing both the goal and the solution space in the same instantly.

The solution: the Thresher, is seen on entry and stands out as the only moving element, drawing the player's attention. Pulling the lever stops the thresher briefly before it turns back on, clearly communicating the cause and effect. This gives the player time to understand the mechanics before they can proceed.

To start the process, I gathered a lot of references. Since the level is inspired by a real location, I used online maps to get references of the biome and height map of the location. I also looked up floor plans, videos documentation I could find for inspiration.

I created a color palette that I combined with concept art and brutalistic architecture references to create my own futuristic spin on this existing area.

Factory Breakdown

This Puzzle is placed right after the player's first enemy encounter, slowing the pace while keeping the tension.

This Puzzle is in two parts:


The first part acts as onboarding, where the player uses a box to get on the truck. The Truck is loaded with similar boxes with some rough decals on them, in order to reinforce the interaction and build visual consistency.


A one way drop then commits the player in to the main puzzle area.

Similar to the earlier farm puzzle, the original exit is blocked, but the new exit is seen in the same frame: Similar boxes that were packed on the truck outside.

The solution is foreshadowed through the gantry crane, which is hoisting a familiar box. Although this is the last time this action is used: releasing it via a lever reinforces a previously learned behavior.

To subvert expectations, dropping the box creates a loud impact that alerts an anemy on the balcony above, introducing pressure to an otherwise predictable interaction.

The central area of the yard provides cover, while the box drop and exit zones are exposed. This is designed for the player to use the box as a movable cover, or retreat to the safe middle zone at the cost of time.

To start the process, I gathered a lot of references. Since the level is inspired by a real location, I used online maps to get references of the biome and height map of the location. I also looked up floor plans, videos documentation I could find for inspiration.

I created a color palette that I combined with concept art and brutalistic architecture references to create my own futuristic spin on this existing area.

Production

Plan Level

To get started I wrote a text document where I outlined the level from start to finish. This served as a foundation in the sketch & blockout stages.

I didn't follow it completely, but having a step-by-step breakdown of events and key questions helped me erase unnecessary spaces and stick to core gamapley beats. I then iterate on the design directly in the editor.

To get started I wrote a text document where I outlined the level from start to finish. This served as a foundation in the sketch & blockout stages.

I didn't follow it completely, but having a step-by-step breakdown of events and key questions helped me erase unnecessary spaces and stick to core gamapley beats. I then iterate on the design directly in the editor.

Plan Level

To get started I wrote a text document where I outlined the level from start to finish. This served as a foundation in the sketch & blockout stages.

I didn't follow it completely, but having a step-by-step breakdown of events and key questions helped me erase unnecessary spaces and stick to core gamapley beats. I then iterate on the design directly in the editor.

References & Hex Codes

I wanted my level to have a similar look as a Tarsier Studios game. To support this, I collected references for each area, along with hex codes for the base colors palette.
The hex codes gave me a great framework when adding assets and materials in the editor,helping me mainting a cohesive visual identity throughout the level.

References & Hex Codes

I wanted each space in the level to feel purposeful and believable, as if the world exists with or without the player. To achieve this, I designed areas with clear functions, such as control panels for the conveyor belt, a locker room, transports parking garage and the weapon generator.

Each space was given its own identity, and the layout of the environment was shaped to support that specific purpose

Paper Design

To stay within scope, I chose to map the level with a bubble diagram, placing spaces in a rough layout of the overall map. The camera's position was considered from the start, with the player view from the bottom of the page.

The layout helped me visualize the player flow instead of details.

My initial plan was two puzzles and two enemy encounters. Due to scope, the forest encounter was scrapped in the blockout phase. Although it could have added a high intensity beat, it was not manageable within the timeframe.

Instead, I merged the second enemy encounter with the second puzzle and kept the forest as a slow pacing beat leading into the battlefield.

Paper Design

To stay within scope, I chose to map the level with a bubble diagram, placing spaces in a rough layout of the overall map. The camera's position was considered from the start, with the player view from the bottom of the page.

The layout helped me visualize the player flow instead of details.

My initial plan was two puzzles and two enemy encounters. Due to scope, the forest encounter was scrapped in the blockout phase. Although it could have added a high intensity beat, it was not manageable within the timeframe.

Instead, I merged the second enemy encounter with the second puzzle and kept the forest as a slow pacing beat leading into the battlefield.

Prototyping

After sketching a top-down bubble diagram, I moved into the editor and set up the fixed camera and camera triggers. This let me manually control what would be visible on the player's screen. I also defined the player metrics, limiting the jump to 1m high, establishing a strong frame to base my traversal puzzles after.

I then set up Ultra Dynamic Sky, Post-process and lighting to test visibility and silhouettes readability in the fog.
This helped me determine distances to short-term goals

To start the process, I gathered a lot of references. Since the level is inspired by a real location, I used online maps to get references of the biome and height map of the location. I also looked up floor plans, videos documentation I could find for inspiration.

I created a color palette that I combined with concept art and brutalistic architecture references to create my own futuristic spin on this existing area.

Prototyping

After sketching a top-down bubble diagram, I moved into the editor and set up the fixed camera and camera triggers. This let me manually control what would be visible on the player's screen. I also defined the player metrics, limiting the jump to 1m high, establishing a strong frame to base my traversal puzzles after.

I then set up Ultra Dynamic Sky, Post-process and lighting to test visibility and silhouettes readability in the fog.
This helped me determine distances to short-term goals

Living world

The spaces between gameplay beats lacked activity.

To adress this, I introduced movement and foreshadowing of upcoming events, making the world feel more alive while maintaining the eerie tone.

I used this approach throughout the level to create a cohesive, diegetic experience driven by visual storytelling.

Living world

The spaces between gameplay beats lacked activity.

To adress this, I introduced movement and foreshadowing of upcoming events, making the world feel more alive while maintaining the eerie tone.

I used this approach throughout the level to create a cohesive, diegetic experience driven by visual storytelling.

Opimization

I build the whole level with meshes, including the landscape. This allowed me to iterate quickly while blocking out.

The problem was that as the number of meshes increase, performance began to drop. To address this, I converted large sections of the landscape into editable level intances, reducing the number of actors while maintaining flexibility.

I used the same approach for finished structures as the barn and factory, improving both my workflow and efficiency.

This allowed me to maintain performance while iterating on large-scale changes.

Opimization

I build the whole level with meshes, including the landscape. This allowed me to iterate quickly while blocking out.

The problem was that as the number of meshes increase, performance began to drop. To address this, I converted large sections of the landscape into editable level intances, reducing the number of actors while maintaining flexibility.

I used the same approach for finished structures as the barn and factory, improving both my workflow and efficiency.

This allowed me to maintain performance while iterating on large-scale changes.

Player Camera Adjustments

Since the player is always moving from left to right, it is important that the player can see what is ahead.
During playtesting the player continously ended up on the right side of the screen. This was adjusted by tweaking all the values in each area of camera triggers so that the player stayed in the middle or on the left side of the screen.

Player Camera Adjustments

Since the player is always moving from left to right, it is important that the player can see what is ahead.
During playtesting the player continously ended up on the right side of the screen. This was adjusted by tweaking all the values in each area of camera triggers so that the player stayed in the middle or on the left side of the screen.

Spotlight Enemy

I chose to create a static enemy, both for scope and it was fitting for the tone of level that I was creating.


The enemy is a human holding the flashlight. While it has a skeletal mesh, it is hard to read when they are standing still, as the player often gets blinded by the light.
This could have been improved by scaling the mesh or by placing the enemy closer to the playable area.

However, I chose to keep it as-is, because it supports the eerie tone of the level: The player is seen, but can barely see the enemy.

The enemy is revealed more clearly in the following chase sequence.

To start the process, I gathered a lot of references. Since the level is inspired by a real location, I used online maps to get references of the biome and height map of the location. I also looked up floor plans, videos documentation I could find for inspiration.

I created a color palette that I combined with concept art and brutalistic architecture references to create my own futuristic spin on this existing area.

Chase Enemy

The chase sequence is triggered by the player activating a sound trap: a string with empty cans attached to it. Alerting an enemy that runs out from the woods and chase the player along th tracks. The sequence ends with the player hiding under a rock to avoid being spotted by the enemy.


To stay within my scope, I reused the existing BP_Enemy instead of creating a new enemy type. I disable the light panner to make the light static, then move the enemy along a spline following the player path.

When the player jumps down from the tracks, a trigger swaps the enemy back to the original panning behavior using a level sequence.

To start the process, I gathered a lot of references. Since the level is inspired by a real location, I used online maps to get references of the biome and height map of the location. I also looked up floor plans, videos documentation I could find for inspiration.

I created a color palette that I combined with concept art and brutalistic architecture references to create my own futuristic spin on this existing area.

Reflections

Overall, I'm happy with the result. If I were to revisit the level, I would refine the onboarding of the core mechanics. However, that was not the main intention, the level is designed with focus on pacing, composition and readability.

As mentioned under "Paper Design", the level would benefit from a peak tension moment that introduces more enemy pressure in the forest before finally slowing down on the battlefield. I believe this would have created a memorable final beat.

I'm drawn to narrative-driven experiences that evoke emotion, and this project was an exploration of that without dialogue, sound or NPC interactions.

Designing a purely visual experience that still creates emotional impact was challenging with the timeframe that I had, but it was a very rewarding process.