top of page
SICARIA | 2024

pre-production

production

20 weeks

39 devs

Game Design

Technical Design

UI Design

Responsibilities

  • Designed and implemented the AI visual detection system

  • Designed and implemented the UI for visual detection

  • Created concepts for the player HUD

  • Created wireframes for the menu's

  • Led playtest sessions and analyzed relevant data

uelogo_edited.png

Unreal Engine 5

Blueprints

image_edited.png

Released on Steam

PROJECT OVERVIEW

Game summary

​Step into the shadows of Antwerp and shapes events during this crucial moment.

Sicaria is a third-person stealth game set in ​Antwerp during the Eighty Years' War. You play as an engineer, a person capable of overcoming any challenge through your own cleverness. Because of the war you lost an arm, but you were able to build a mechanical one.

 

Game details

  • Genre

  • Player support

  • Platforms

  • Release date

Stealth, action

Single player

Steam

April 12, 2024 (early access)

GAME OVERVIEW

Research and planning

Before jumping into engine, I took time to research the ins- and outs of the visual detection systems in other games, such as Splinter Cell and Alien: Isolation. For the detection boxes, I used human metrics to create the first iteration. The boxes were made to be 220 degrees, 120 of those being what we consider ''direct line of sight''. The idea was to give the enemies realistic vision (i.e being able to see out of the corners of one's eyes also known as peripheral view), but keep it fair and scaled. 

Inspired by Splinter Cell, I went with a bone detection system to register when a player is spotted. I would set a number of bones in the players skeletal mesh as ''detectable'' to which the enemy could trace lines while the player is in the detection boxes. When a bone has been detected, the system would check pre-determined criteria to see if the amount of bones detected would register the player as seen when taking into account the distance between the player and enemy. This way, we would be able to differentiate that the enemy would understand they are seeing someone when seeing a leg right next to them, compared to seeing a leg 20 meters away. 

 

On the numbers side, I brainstormed the elements that affect detection, which I called ''multipliers''. These would be things such as movement and being close to the enemy increasing detection, being crouched lowering the increase of detection, etc.

 

Responsibilties

  • Researched visual detection systems in other games

  • Concepted, modelled and implemented the visual detection boxes

  • Planned the visual detection system

Modelled detection cones in Unreal Engine 5

References

VISUAL DETECTION SYSTEM

System

Once the system was planned out, it was time to hop into engine. The system itself was set up within a Blueprint component so that we could easily attach it to the enemy actors and keep all functionality related to visual detection in one place. Once the base system was done, I went through numerous iterations to balance the numbers, reduce any unneccessary complexity, clean up the Blueprints, and to help optimize the system. Once the system was ready to test, I conducted a number of playtests and gathered feedback.

 

A major point of feedback was that the players felt that they were detected too fast. Feedback like this was important to tackle, as I wanted visual detection to feel like a challenge, not a punishment. On further inspection it appeared that the players did actually agree that they should have been spotted when they were, but that they felt like they had little time to react. Instead of adjusting the numbers, I added a ''forgiveness window'', so that if the player had been spotted for the first time, no matter how close to the enemy, they would have a 1-second window to re-adjust their positioning to avoid detection. This additionally helped the cases where the player may have accidentally uncrouched at the wrong time, or made a slight misstep. Once the ''forgiveness window'' was implemented and tested, it was noted that it had greatly improved the players experience, 

Responsibilties

  • Created and implemented the system in Blueprints​

  • Iterated on and optimized the system

  • Iterated on the connection between visual detection and the AI state handling

  • Add quality of life features such as the ''forgiveness window''

As I sometimes had to visit our enemy AI base blueprint, I took it upon myself to clean up and optimize the Blueprint as it was not well taken care of 

Base functionality blueprint

UI 

An important part of visual detection in games, especially stealth games, is clarity and feedback. Feedback is often given to the player through UI, audio, visual effects, etc. As the designer of the visual detection system, I took on the task to design and implement the UI. The visual detection UI consisted of 2 UI components, one being the on-screen UI (the player sees the enemy that sees them), and the other being off-screen UI (the player does not see the enemy that sees them). 

For the on-screen UI I took inspiration from A Plague Tale Requiem. We wanted the player to not only know they are being detected, but what the enemy's current behaviour is so that they could act accordingly. I chose to communicate this with simple icons and colors that would indicate that the players presence is being noticed, that they are being searched for, and that they are under attack.

For the off-screen detection I took inspiration from Metal Gear Solid V: Ground Zeroes. Some games handle off-screen UI by displaying icons around the edges of the screen, pointing towards objectives. A problem with this idea is that players can be overwhelmed by large amounts of small UI pieces moving around, and in the cases of visual detection, it becomes hard for the player to understand where their safe space is. Games like Metal Gear Solid center detection information around the character model, so that it is located within the general window of the players focus

Responsibilties

  • Researched visual detection UI in other games

  • Created the concept for the on-screen and off-screen UI for visual detection

  • Created and implemented UI

Icon designs 

Showcase of the visual detection UI

Research and planning

The UI artist and I researched and established the designs of the menu's and HUD.

 

On the menu side we discussed how we could keep the menu's simple but aesthetically pleasing, and how we would use different types of fonts where necessary (for example, having non-decorative fonts for larger parts of text such as subtitles and the settings menu).

 

On the HUD side, we carefully considered things such as positioning, visibility (for example, removing the stamina bar from the screen when the player has full stamina to focus on immersion), contrast against the game screen, etc. Together we found a balance between design and art while keeping it all accessible, and we produced concepts which were later implemented by me and others.

 

Responsibilties

  • Researched menu's and relevant HUD elements

  • Created wireframes for the menu's

  • Helped concept the HUD elements

  • Implemented the HUD elements

Showcase of the current menu's and HUD

Wireframes of the menu's