💻 computer science

AEDHunter: Investigating AED Retrieval in the Real World via Gamified Mobile Interaction and Sensing

The paper presents AEDHunter, a gamified mobile application that leverages smartphone sensing and Bluetooth verification to facilitate in-situ training, demonstrating that repeated practice significantly improves users' AED retrieval times and confidence while providing behavioral metrics to quantify learning progress.

Original authors: Helinyi Peng, Akihito Taya, Yuuki Nishiyama, Kaoru Sezaki

Published 2026-03-03

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Helinyi Peng, Akihito Taya, Yuuki Nishiyama, Kaoru Sezaki

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where a heart stops beating outside a hospital. In those critical minutes, every second counts. An Automated External Defibrillator (AED) is like a "life-saving reset button" that can restart a heart, but here's the problem: most people don't know where the buttons are hidden.

Even if an AED is right down the hall, panic and confusion can make it feel miles away. This is where AEDHunter comes in. Think of it not as a boring safety manual, but as a real-life treasure hunt designed to train your brain to find these life-saving devices without thinking twice.

Here is the story of how this app works, explained simply:

🎮 The Game: "Treasure Hunt for Life"

Imagine you are playing a video game, but instead of a screen, the whole campus or office building is your level.

The Mission: The app tells you, "Someone has collapsed! Find the nearest AED immediately!"
The Timer: A clock starts ticking, and a "survival bar" drops rapidly, showing you how every second lost lowers the victim's chance of living. This creates a sense of urgency, just like in a real emergency.
The Hunt: You have to physically walk to the building, find the right floor, and locate the specific cabinet.
The Twist: You do this over and over again. Just like practicing a sport, the more you play, the better you get.

🧠 How It Trains Your Brain (The "Mental Map" Analogy)

Before playing, most people are like tourists in a new city: they rely on a paper map (or a GPS app) to get from Point A to Point B. If the map is wrong or they panic, they get lost.

AEDHunter turns you into a local.
By running the same "treasure hunt" route repeatedly, your brain stops needing the map. You start building a mental map. You stop thinking, "Which way is the elevator?" and start thinking, "I know this building; the AED is on the second floor, left of the lobby."

The study found that after just a few practice sessions, people found the AEDs significantly faster. They stopped hesitating and started moving with purpose.

🕵️‍♂️ The Secret Spy: How the App "Watches" You

The app is smart. It uses your phone's sensors (like the accelerometer that knows when you're walking) to act like a detective.

The "Hesitation Signal": The app can tell the difference between you walking confidently and you stopping to think, "Wait, is this the right building?" or "Where is the door?"
The Result: As people practiced, the app noticed they stopped "stopping." They didn't pause to think as much. This "hesitation time" dropped by about 40%. This proves they weren't just running faster; they were thinking less because they knew where to go.

📊 What Happened in the Real World?

The researchers tested this on university students. Here is what they found:

Speed: The time it took to find an AED dropped from an average of 2 minutes and 12 seconds down to 1 minute and 37 seconds. In a heart attack, that extra 35 seconds could be the difference between life and death.
Confidence: Before the game, many students were unsure if they could find an AED. After playing, they felt much more confident they could do it in a real emergency.
Awareness: Even when they weren't playing the game, people started noticing AEDs in their daily lives. They went from being "blind" to them to actively looking for them.

🌟 The Big Takeaway

We often think safety training is about reading a brochure or watching a video. But AEDHunter shows that practice makes perfect, even for emergencies.

By turning a scary, high-stakes task into a fun, repeatable game, the app helps us build muscle memory for our brains. It transforms a stranger in a crowd into a confident rescuer who knows exactly where to run when seconds matter most.

In short: AEDHunter doesn't just teach you where the life-saving button is; it trains your brain to find it automatically, so when a real emergency happens, you don't have to think—you just act.

1. Problem Statement

Out-of-hospital cardiac arrest (OHCA) has low survival rates, largely due to delays in defibrillation. While Automated External Defibrillators (AEDs) are increasingly available, their utilization remains critically low (e.g., <2% in some countries) due to:

Low Public Awareness: Most people do not know the locations of nearby AEDs.
Navigation Challenges: Existing solutions (static 2D maps) fail in urgent, high-stress scenarios, particularly indoors where GPS is weak and building layouts are complex.
Lack of Empirical Validation: Few studies objectively measure whether gamified awareness interventions actually improve real-world retrieval speed and behavior.
Behavioral Gaps: There is a lack of fine-grained data on how people search for AEDs (e.g., hesitation, decision points) to inform better interventions.

2. Methodology

System Design: AEDHunter

The authors developed AEDHunter, a gamified, location-based mobile application (iOS) designed to transform AED retrieval into a repeatable, low-stakes practice experience.

Core Mechanics: Users initiate "hunting" sessions where they must locate a specific AED within a simulated emergency context (e.g., "Someone collapsed, get an AED now").
Gamification Elements:
- Urgency Simulation: Real-time display of elapsed time and a dynamically decreasing "survival probability" curve (based on the exponential decay of survival rates over time).
- Feedback: Immediate post-session surveys and performance metrics.
- Modes: "Routine Game Sessions" (spontaneous practice) and "Exam Sessions" (structured pre/post assessments).
Hardware & Sensing:
- Sensors: Utilizes smartphone sensors (GPS, Wi-Fi BSSID, accelerometer, gyroscope, barometer, magnetometer) and Bluetooth Low Energy (BLE) beacons placed near AEDs for arrival verification.
- Data Collection: Logs movement trajectories, sensor data at 100Hz, and Wi-Fi signals for indoor entry detection.

Experimental Setup

Participants: 20 university students across two campuses (Campus I and II).
Groups: Participants were assigned to a Map Group (received map guidance) or a No-Map Group (relied on distance/direction cues only) to test the impact of navigation aids on spatial memory formation.
Procedure:
1. Pre-Exam: Baseline retrieval of a specific AED.
2. Training Phase: 3 weeks of routine game sessions (repeated practice).
3. Post-Exam I: Retrieval of the same AED to measure learning.
4. Post-Exam II: Retrieval of a new, unfamiliar AED to measure knowledge transfer.
5. Follow-up: A subset of participants returned 6–7 months later to test long-term retention.

Behavioral Analytics & Classification

Phase Segmentation: Trips were segmented into three phases using sensor data:
1. Preparation: Stationary planning (optional).
2. Building Search: Walking to the target building.
3. Indoor AED Search: Locating the device inside.
Exploratory Pauses: A lightweight Support Vector Machine (SVM) classifier (trained on accelerometer and angular velocity data) distinguished between "moving" and "exploratory pausing" (hesitation/uncertainty).
- Metric: The reduction in pause duration ( $\Delta D_P$ ) served as a behavioral learning signal.

3. Key Contributions

Gamified In-Situ Training: Demonstrated that a mobile app can effectively turn AED discovery into an engaging, repeatable practice loop that embeds spatial memory.
Sensor-Driven Behavioral Insights: Introduced a novel method to quantify "hesitation" via exploratory pauses using smartphone sensors, providing a granular view of the retrieval process beyond simple completion time.
Empirical Validation: Provided objective, real-world evidence that repeated gamified practice significantly reduces retrieval time and hesitation, moving beyond subjective self-reports.
Design Framework: Established a framework for "just-in-time" adaptive interventions and validated the efficacy of low-cost Bluetooth verification for indoor tracking.

4. Key Results

Quantitative Performance

Retrieval Time Reduction: The median total retrieval time decreased significantly from 132.7 seconds (Pre-exam) to 97.3 seconds (Post-exam I), a reduction of ~35%.
Relative Improvement: The median relative retrieval-duration reduction ( $\Delta D_T$ ) was 0.39 (39% faster), with 85% of participants showing improvement.
Hesitation Reduction: The median relative pause-duration reduction ( $\Delta D_P$ ) was 0.40, indicating a 40% decrease in exploratory pauses. This suggests users became more confident and less uncertain during the search.
Phase-Specific Gains:
- Building Search: Showed the most significant time reduction (median -27.1s), indicating improved route planning and building identification.
- Preparation: The "planning" phase largely disappeared in the Post-exam, suggesting users had internalized the location and no longer needed to stop and plan.
Group Comparison: Both Map and No-Map groups improved significantly. Interestingly, the No-Map group showed greater improvement in the "Building Search" phase, suggesting that relying on memory (without maps) may foster stronger spatial learning for the overall route.

Subjective & Behavioral Outcomes

Confidence & Familiarity: Self-reported familiarity with AED locations increased dramatically (median score rose from 1.0 to 5.0). Confidence in retrieving an AED also increased significantly.
Awareness Shift: Participants shifted from passive unawareness to active scanning. Post-study, a significant portion reported actively looking for AEDs in their daily environment.
Long-term Retention: In the 6–7 month follow-up, retrieval times for the practiced AED remained faster than the initial baseline, though slightly slower than the immediate post-exam, indicating partial retention of spatial memory.

5. Significance and Implications

Public Safety Impact: By reducing retrieval time by ~35 seconds, the system addresses the critical "golden hour" of cardiac arrest, where every minute of delay reduces survival by 7–10%.
Scalable Training: The "bring-your-own-device" approach requires no dedicated hardware infrastructure beyond low-cost BLE beacons, making it scalable for universities, offices, and public spaces.
Design Insights:
- Dynamic Cues: Time and survival probability displays were more motivating than static figures.
- Map vs. No-Map: While maps help initially, removing them may accelerate the development of robust internal spatial maps.
- Environmental Design: The study highlights the need for better physical signage (distinctive colors, lighting) to complement digital training, as uniform building interiors hindered visual recognition.
Future Directions: The paper suggests integrating real-time classifiers for "just-in-time" hints (e.g., detecting a pause and suggesting a direction) and expanding studies to diverse populations and high-stress simulated environments.

In conclusion, AEDHunter proves that gamified, sensor-enabled mobile training is a viable and effective strategy for improving bystander preparedness, transforming AED retrieval from a confusing, high-stress task into a practiced, efficient skill.