UMC Utrecht
In Brief
The Utrecht Brain-Computer Interface (BCI) Lab, part of the UMC Utrecht Brain Center, is a multidisciplinary research group dedicated to translating cutting-edge neuroscience into transformative solutions for individuals with severe motor impairments. Addressing the limitations of the previous BCI training game, ‘Whack-a-Mole’ (the ‘click task’), I created Dungeonfall, a 2D grid-based game concept designed to increase user motivation, improve data labelling, and train binary click interaction.
Role
UX Designer
UI Designer
Visual Designer
Interaction Designer
Timeline
RESEARCH
The Whack-a-Mole: a Training Game
Context
For individuals with locked-in syndrome (LIS) caused by late-stage amyotrophic lateral sclerosis (ALS), regaining control over intentional movements is a significant challenge.
The Whack-a-Mole ('click task') Brain-Computer Interface (BCI) training game, designed for a laptop-sized screen, was used in the Utrecht Neural Prothesis (UNP) system during a research project by the UMC Utrecht BCI Lab in 2021.
Its purpose was to help users improve binary brain-controlled selections, the ability to make intentional "clicks" using their neural signals.
How the Training Game Works
Click Threshold
A click is triggered when the brain signal surpasses a set threshold for at least 1 second, requiring precision and sustained effort.
Scanning Mechanism
The grid highlights rows and columns sequentially at a scan rate of 2 seconds per row or column. Users have to click when the box highlights a mole.
Feedback Categories
The training game evaluates user performance through four feedback types:
True Positives (TPs): correct clicks during target intervals.
False Positives (FPs): unintended clicks outside valid target intervals.
False Negatives (FNs): missed clicks during intended intervals.
True Negatives (TNs): non-click moments when no target was present.
These categories allowed researchers to systematically analyse the system's accuracy and identify areas for improvement.
The Research Key Findings
Problem
While the mechanics formed the foundation of the training game, real-world interaction highlighted four key challenges impacting accuracy and user experience.
IDENTIFY
The Problems That Limited the Training Experience
Based on research findings and team discussions, I highlighted four urgent design interventions:
Minimise Unintended Actions
False Positives (FPs) showed a high rate of unintended clicks.
Increase User Motivation
Lost motivation and engagement due to repetitive mechanisms.
Reduce Cognitive Barriers
Users struggled to time the click due to the scan rate.
Improve Game Adaptability
A one-size-fit approach proved to be ineffective.
DEFINE
Understanding Our Users: Who Are We Designing For?
Our users are primarily middle-aged men and women with late-stage ALS and experiencing locked-in syndrome (LIS). This condition leaves them unable to move but with their cognitive abilities intact.
To better understand their challenges and needs, I categorised key insights into two areas:
IDEATE
Clearing the Fog — Framing the Problem
At this point in the project, I formulated problem statements to frame each issue. To turn these challenges into opportunities, I used 'How Might We?' statements to guide the ideation of solutions for a new training game concept.
Designing for Impact—Targeting Key Challenges
With these questions as a foundation, I identified the most critical design challenges to tackle, ensuring the training game would be more precise, motivating, and adaptable.
Up to this point, I worked closely with my team, sharing feedback and ensuring our technical requirements aligned with user needs.
But I also wanted to ensure we focused on what mattered, balancing priorities with what was feasible from an engineering standpoint in the short term and what could be tackled later.
Brainstorming and Sketching
With these newly framed challenges in mind, I focused on two key goals: generating many ideas and many alternatives. What mattered here was brainstorming as many ideas as possible without judgment and mind limitations.
How I did it:
My team and I brainstormed and sketched as many ideas as possible. I have always preferred writing, drawing in a pen-and-paper notebook, and later working on my laptop.
While doing that, I also checked whether these ideas fit the project requirements and the results we found in the previous steps. This procedure helps me later when I have to narrow down and finalise the ideas.
Finalise Ideas — Initial Concept: a Moving Circle
Concept decision
We decided to test one idea from the sketching and brainstorming session: making a circle move around a grid. By clicking an invisible tile, the circle can move in all directions.
After testing it, we identified a few issues. Despite the simplicity, there was still a risk of inaccurate clicks, and too much user freedom could lead to misinterpretation by the UNP system.
We, therefore, refined it into a step-by-step scanning process, highlighting the final goal location before clicking it. By chunking user actions into smaller, manageable tasks, we aimed to:
1
Training binary click interaction.
2
Improving data labelling accuracy.
3
Enhancing user experience and motivation.
Tackling Monotony: Introducing Levels and Gameplay Variety
During my ideation process, I brainstormed and researched ways to make the game dynamic while keeping its core mechanics simple, ensuring user motivation and consistent interaction.
Levels and exploration game-based emerged as one way to vary the game while maintaining the structured training approach.
Why Levels?
Variety Within Constraints
Dividing the game into levels lets us add new features, challenges, and environments while keeping the core binary click mechanics intact.
Data Collection Opportunities
The multi-level design ensures repeated interactions, allowing researchers to gather more data while maintaining user interest.
Structured Progression
Levels provide a clear sense of progress and achievement, keeping users engaged and motivated to continue playing.
Building Challenges into the Game
With levels in mind, I envisioned a main character navigating challenges like defeating enemies or overcoming obstacles to complete levels.
These challenges train binary clicks, offer varied interactions, and enhance motivation and engagement.
Rewards and Penalties: Motivating Interaction
Drawing inspiration from gaming design principles, I proposed integrating rewards and penalties to sustain engagement:
Rewards
Overcoming challenges unlocks features like character customisation or collectables, providing a sense of ownership and accomplishment.
Penalties
Failing tasks may lead to mild setbacks, like reattempting a section, creating a balance between challenge and accessibility.
Adding Mini-Games: Games Within the Game
I proposed incorporating mini-games as part of the level structure to diversify the gameplay further. These mini-games would serve to:
1
Train specific tasks in focused scenarios, such as timing clicks or navigating paths.
2
Add an element of surprise and novelty, breaking the monotony of core gameplay mechanics.
3
Provide rewards, increasing motivation and engagement.
DESIGN
Game and Design Style
We decided to design a prototype of a single-player RPG (role-playing game) to explore user motivation, engagement, and training binary clicks through a progression-based quest.
Players assume the role of a pixel art character who navigates various environments and solves mini-games to restore light.
The choice of pixel art enhances the game's accessibility by providing a clean, simplified view. This simplicity is functional, as it strips down unnecessary details, making key interactive elements easy to identify and engage with.
Color Palette
I used complementary colour schemes to create a balanced yet engaging game environment. The earthy tones with warm accents, like yellow-orange, stand against the dim, gloomy blue-purple tones.
0D2B45
203C56
554E69
8D697A
D08159
FFC183
FFD4A3
FFECD6
000000
These contrasts guide users’ focus toward interactive items, ensuring visual clarity and enhancing gameplay accessibility.
Art Direction
Adopting a modular approach, I used a central square-shaped building block to create complete scenarios, which suits the pixel art style.
UI Elements
I designed UI Elements using the modular approach I used for the different components of my game design.
There are three starting-point from where the scanning can begin:
1
It can start based on the UNP system's prediction of where the target can be (based on the past target's position).
2
It can start from the right side of the character's current position.
3
It can adapt based on the system's understanding of the user's behaviour and goals (Advanced mode).
Why Three Different Starting Point?
Given the uncertainty of the improved UNP system's accuracy in assessing user data intentions, these concepts explored various entry points for the scanning process, aiming to create a flexible approach for future scanning methods.
Example of scanning process starting from the right side of the character:
This step-by-step approach allows users to make precise decisions gradually, improving accuracy and familiarity with the controls.
The segmented movements reduce misinterpreted signals, ensuring intentional and precise directional choices.
Finally, the character itself cannot be clicked, as the system scans only the grid tiles, not the player.
What if I Make a Mistake?
If the system infers the selection as wrong, once confirmed, it turns red and disappears after a couple of seconds, letting the user start a new scanning process from scratch.
Three Playable Mode Settings
To provide researchers with precise data on player decision-making, the game includes three primary modes of play:
1
Single-path mode.
2
Advanced-path mode.
3
Hybrid-path mode.
Single-Path Mode
This mode is initially used to ease players into the game by reducing cognitive load and minimising distractions. It focuses on precision and data accuracy.
Adavanced-Path Mode
In Advanced-path mode, the player has complete freedom of movement and decision-making. This mode focuses on boosting motivation and engagement, allowing the user to try to reach the targets quickly.
This mode is ideal for tracking player preferences and strategies in a less controlled environment, allowing researchers to observe decision-making patterns in a fully open setting.
Hybrid-Path Mode
The Hybrid-path mode combines elements of the Single-path and Advanced-path modes, providing the player with alternating levels of constraint and choice.
This mode allows gradual adaptation, bridging the transition from highly structured gameplay to open exploration.
Obstacles
The obstacle functionalities are meant to train and reinforce the binary click and the grid system interaction.
Clickable obstacles
Enemies Fighting Mechanism
Fighting the different enemies in the training game allows us to reinforce both binary and timed clicking.
Each enemy has an energy bar indicating their health (100% complete at the start). In combat, the user has three clicks to defeat the enemy:
1
The user selects the enemy via the scan system.
2
The user clicks on the enemy three times in a row, with each click reducing the enemy's health by 33%.
3
On the third click, the enemy is defeated.
Collectable Items
Items serves to training binary click and enhance the user motivation.
Collectable coins interaction
Collectable key interaction
Games Within the Game: Mini-Challenges for Enhanced Training
The game includes a series of mini-games designed to challenge the user, provide variety, and add layers of engagement.
Each mini-game requires problem-solving, quick reflexes, or strategic thinking and rewards players with new customisations and characters.
They are meant to reinforce gameplay mechanics, train specific skills like time-based clicking or precision, and increase engagement and motivation by adding variety to the experience.
Examples of Mini-Games
The training game offers two different mini-game modalities:
1
A specific part of the map (room) where the user has to complete a challenge or a task before moving on to the next section of the game.
2
The user can complete a mini-game by selecting a specific item that will open the said mini-game.
Penalties and Rewards
By balancing penalties and rewards, we create a gameplay concept loop that challenges users while motivating them to continue playing. This system also increases the volume of data we can collect.
Rewards system
Rewards are structured to enhance player engagement and sense of accomplishment, with unique items that provide narrative progression, customisation, and gameplay advantages.
Penalties system
The penalties system adds tension and encourages thoughtful gameplay, challenging players to balance caution and efficiency.
Initial Menu Features
When users launch the game, an initial menu offers options to customise their gameplay experience. These options include game modes, time adjustments, and starting point selections to suit different user preferences.
Game Difficulties Settings
1
Death Setting:
2
Time Setting:
The game accommodates users' skill levels and pacing preferences, allowing players to shape their experience, making it more inclusive and tailored to individual comfort levels.
TESTING
Time to Test the New Concept Training Game Usability
I conducted a moderated usability test with five users without late-stage ALS. to test the game prototype's usability
This was because recruiting actual users with ALS for research is a lengthy and highly selective process.
Finding a single participant for a whole research project can take up to four months, and collecting sufficient interaction data typically requires a full year.
But why a moderated usability testing?
1
I aimed to build relationships with my participants and help them open up to share feedback.
2
I wanted to have the chance to talk to them after finishing tasks and going more into detail.
3
The goal was to maximise firsthand observations, offer participants technical support, and clarify potential misunderstandings.
Design Impact
From Data to Insights
Insights
Users found some inconsistencies with the menu settings.
Users found the grid system a bit tricky to use without prior knowledge. Once they understood it, it was easy to use.
The enemy animations helped to understand the dynamics and convey that it was required to hit it again.
The relationship between the closed door and the key hidden inside the treasure box could have been clearer and more straightforward.
Resulting Revisions
1
I changed the "Play" label to "Game mode", making it more evident that there are features to discover and choose from.
2
I did the same thing for the "Start game" label to "New game", again making it more explicit where to click when a user wants to start a new game from the beginning.
3
Finally, the new "Resume game" label for the old "Resume checkpoint" saves reading time and helps clarity.
In the usability test, most participants asked for more straightforward and intuitive button labels.
The initial design challenged users to find the key without any direction but to encourage binary clicks and gather more data; we aimed for a more intuitive approach.
Usability feedback showed the need to strengthen the door-key relationship, leading us to add a visual cue for immediate correlation.
NEXT STEPS
UX is Never Done
While significant progress has been made, more must be done. Moving forward, the priorities will be:
Implementing More Onboarding Solutions
During usability testing, we found that users needed a more straightforward introduction to the gameplay. No instructions were provided since we aimed to test its intuitiveness without guidance.
Simplify Gameplay
Providing a gradual introduction to the game mechanics and presenting new elements one at a time.
Tutorials
Easing the learning curve with guided tutorials and practice areas.
Designing Customisation Features
The next step is adding customisation features to enhance engagement, preserving core mechanics to boost replayability and motivation.
Iterating Through More Usability Testing
To refine the game experience, additional usability testing with more participants will be conducted to validate improvements and identify further refinements.
Enhancing Data Collection for Deeper Insights
To gain deeper insights into usability and engagement, we will explore implicit data collection, task success rates, and performance trends—tracking improvements, errors, and consistency over time.
REFLECTIONS
Four Key Milestones
In designing the new training game, I focused on making interactions more engaging, intuitive, and effective for data collection.