- Visit the Hosted Project: Open your browser and navigate to: http://fallsafe.hellojeffreylee.com:8000
- Clone the Repository:
git clone https://github.com/CheeGuang/FallSafe
- Request the Environment File:
- Contact the administrator (Jeffrey, @CheeGuang on Telegram) to obtain the
.env
file and place it in the root directory.
- Run Kubernetes Deployment:
deploy-k8s.bat
- Perform Port Forwarding:
port-forward-all.bat
- Access the Application:
- Open your browser and navigate to
http://localhost:8080
.
This project is developed where our key stakeholder is Lions Befrienders. The problem statement we are addressing is:
How might we use Technology to offer an easy and user-friendly way for seniors to perform preliminary self-assessment in the comfort of their own homes, with vulnerable cases being highlighted for further clinical assessment by doctors?
Our solution is a 100% Cloud Native application with Microservice Architecture, allowing seniors to take the Falls Efficacy Scale test and perform physical self-assessments from the comfort of their homes.
FallSafe is designed to support elderly individuals who want to assess their fall risk easily and accurately from the comfort of their home. Our primary user persona, Mrs Tan, represents the key demographics and challenges our solution aims to address.
- Check fall risk at home – No need for clinic visits.
- Understand her body – Gain insights into mobility and balance.
- Visit the doctor conveniently – Seek medical help if the risk is high.
- Quick & accurate self-assessment – Needs a reliable but easy-to-use tool.
- Understanding results – Prefers simple, jargon-free explanations.
- Easy doctor access – Requires a seamless way to book consultations if needed.
By addressing these needs, FallSafe empowers elderly individuals to proactively manage their fall risk, promoting safety, independence, and well-being.
We integrate IoT and AI into our solution to enhance user experience:
- A multi-lingual Text-to-Speech model guides users on how to perform tests.
- The FallSafe Device tracks real-time body movements and sends data to our Cloud application via AWS IoT Core.
- The system processes this data to generate a comprehensive and easy-to-understand result dashboard.
After completing a test, users:
- Receive $10 NTUC Vouchers.
- Can schedule a teleconsultation with a doctor via Calendly, which integrates with Google Calendar and Google Meet.
On the admin side:
- They can track users who have not completed their tests and send reminders.
At FallSafe, protecting user privacy and data security is our top priority. We implement industry-leading security standards to ensure the confidentiality, integrity, and availability of sensitive information. Our security measures include:
- Regulatory Compliance: Aiven Cloud Database adheres to GDPR, HIPAA, ISO 27001, SOC2, ensuring strict data protection and privacy standards.
- Secure Cloud Storage: Utilizes AWS S3 for encrypted data storage.
- Cloud Infrastructure Security: Leveraging Google Cloud and Render, ensuring robust security, reliability, and scalable performance.
- AI and Multilingual Security: OpenAI and Google Translate are integrated with strict access controls, ensuring data privacy while enabling multilingual accessibility.
- Docker & Kubernetes: Implementing containerized security to ensure fault isolation, scalability, and automated recovery, protecting against system vulnerabilities.
- Google Meets: Secure communication channels with end-to-end encryption, safeguarding user interactions.
- SMTP Protocol: Encrypted email transmissions ensure secure delivery of notifications and communications.
- Calendly Integration: Secure scheduling system ensuring protected user interactions with Google Calendar and Google Meet.
- Integrated with Google Translate and OpenAI, allowing users to take this internationally recognised fall risk assessment easily.
- Supports multiple languages, ensuring accessibility for diverse user groups.
- Uses AWS IoT Core to track real-time body movements and assess balance and coordination.
- Provides instant feedback and personalised recommendations for fall prevention.
- Users receive $10 NTUC vouchers upon completing the fall risk assessment with a 6 months cool down.
- Encourages engagement through gamification mechanics, increasing participation rates.
- Enables repeat assessments for continuous fall risk monitoring, ensuring seniors stay proactive in maintaining their health.
- Powered by OpenAI, offering actionable insights based on assessment results.
- Uses Chart.js to present comprehensive charts for a clear visual representation of fall risks.
- Helps seniors track progress over time and make informed health decisions.
- Seamless integration with Calendly, Google Calendar, and Google Meet.
- Enables users to book virtual consultations with doctors for further assessment.
- Ensures timely medical intervention for high-risk users.
- Provides an overview of all seniors' fall risk levels.
- Enables administrators to track test completion rates and identify high-risk cases.
- Supports data-driven decision-making for elderly care organisations.
- Allows admins to send reminders to elderly users to take their fall risk assessments.
- Ensures higher engagement and completion rates for the self-assessment tests.
- Supports SMS and email notifications for convenience.
In total, we conducted 5 Official User Tests with individuals aged 55-85 years old. Through these sessions, we gathered valuable feedback to fine-tune Fall Risk calculations, enhance UI/UX for elderly users, and help them become more independent in their fall risk assessments. These tests also helped identify edge case bugs and improve the representation of results in the Dashboard.
Each test included:
- Evaluation of the Falls Efficacy Scale, Self-Assessment, and Dashboard for fall risk analytics.
- Unmoderated free-roam time to explore the website and interact with the system independently.
FallSafe saw significant improvements with each iteration, ensuring a more user-friendly and accurate experience for seniors.
We plan to integrate FallSafe into Lions Befrienders' Our Kampung Application, enabling users to leverage their mobile phone's gyroscope as a Body Motion Tracker. This reduces the need for separate motion tracking devices, lowering costs and improving scalability.
Additionally, we can tap into the Smart Nation Singapore initiative by leveraging the Mobile Access for Seniors program, which provides a subsidiesd mobile phones starting from $20 and a 2-year mobile plan at $5.10 per month. This will enable seamless digital inclusion, ensuring seniors have access to FallSafe without financial barriers.
Our Kampung is an application by Lions Befrienders (LB) aimed at bridging the digital divide for seniors by providing access to digital services, health tracking, and community engagement.
The Smart Nation Mobile Access for Seniors initiative offers a 2-year mobile plan at $5.10 per month, enabling seniors to stay connected affordably. This initiative ensures digital inclusion and accessibility, making it easier for seniors to engage with technology-driven solutions like FallSafe.
The FallSafe architecture is designed as a Cloud-Native Microservice-Based System to maximize ROI and operational efficiency while ensuring resilience and adaptability. This approach allows the system to react to failures, recover quickly, and minimize business impact.
-
Microservices Architecture
- Each core function (User, Authentication, Admin, Self-Assessment, Falls Efficacy Scale, and OpenAI) is independently containerized using Docker and orchestrated via Kubernetes.
- This enables scalability, fault isolation, and service independence, allowing parts of the system to be updated or recovered without affecting the entire application.
-
Kubernetes & Auto Scaling
- The system is deployed on AWS EC2 instances running Kubernetes, ensuring automated scaling and high availability.
- Load balancing and resource allocation dynamically adjust based on user demand, optimizing costs and improving performance stability.
-
Route 53 Subdomain for Custom Access
- The application is accessible via http://fallsafe.hellojeffreylee.com:8000/, configured using AWS Route 53.
- Custom DNS routing ensures users can reach the application seamlessly, enhancing branding and accessibility.
- The subdomain points to an EC2 instance running Kubernetes, allowing for high availability and scalability.
-
IoT Integration
- FallSafe Device (ESP32S3) connects via AWS IoT Core using MQTT Message Queue, transmitting real-time motion data to the cloud.
- The frontend connects to the application via WebSockets for real-time communication with the FallSafe Device.
- Results is then displayed in real-time on the User Dashboard using Chart.js.
- OpenAI analyzes results, providing actionable insights on how users can maintain and improve their fall risk.
-
Serverless Email & AI Integration
- AWS Lambda and AWS Simple Email Service (SES) handle automated notifications.
- The OpenAI microservice integrates advanced AI models for data analysis and multilingual text to speech.
-
Seamless User Experience with Third-Party Services
- Calendly integration allows users to schedule teleconsultations with Google Calendar and Google Meet.
- Secure authentication and Gmail SMTP ensure smooth user interactions and communication.
-
CI/CD with Docker Hub
- Docker Hub is utilized as part of the CI/CD process, ensuring that EC2 instances can pull the latest container images and deploy updates seamlessly.
- This enhances deployment efficiency, minimizes downtime, and ensures consistency across environments.
-
AWS Security & Monitoring
- The architecture leverages AWS WAF, Shield, GuardDuty, and IAM for robust security enforcement.
- AWS CloudWatch and AWS CloudTrail provide real-time monitoring, logging, and auditing, enabling rapid issue detection and resolution.
- CloudWatch Alarms are configured with SNS notifications, alerting administrators to critical system events, ensuring proactive issue management and minimizing downtime.
- Data is stored on Aiven's Cloud MySQL Database, which offers a 99.99% SLA and complies with ISAE 3000, ISO 27000, GDPR, CCPA, HIPAA, and PCI DSS security standards.
This architecture ensures high availability, flexibility, and resilience, enabling FallSafe to operate efficiently while maintaining a seamless user experience.
The FallSafe system utilizes Kubernetes for efficient container orchestration, ensuring smooth deployment, management, and scaling of microservices.
- Automated Scaling – Kubernetes dynamically adjusts resources based on demand, optimizing system performance.
- High Availability – Ensures fault tolerance by redistributing workloads in case of failures.
- Service Isolation – Each microservice runs in an independent container, reducing dependencies and improving modularity.
- Rolling Updates & CI/CD Integration – Updates are deployed seamlessly with minimal downtime.
- Self-Healing Mechanisms – Kubernetes automatically restarts failed containers and replaces unhealthy nodes.
The FallSafe architecture is designed with cloud scalability in mind, leveraging AWS cloud services to handle fluctuating demand efficiently.
-
Elastic Compute Scaling – AWS EC2 On-Demand instances scale based on workload demands, ensuring cost efficiency and optimal performance without over-provisioning.
-
Serverless Scalability – AWS Lambda handles event-driven processes like email notifications, adapting seamlessly to traffic spikes.
-
Database Scaling – The system uses Aiven Cloud MySQL, which supports automatic replication and read scaling, ensuring high availability and fast query performance.
-
Scalable Storage – AWS S3 provides infinite storage scalability, allowing secure and cost-effective data handling for user-generated content and system logs.
-
IoT Data Processing – AWS IoT Core scales effortlessly to handle increasing motion tracking data from multiple FallSafe Devices, ensuring real-time analytics.
By implementing container orchestration with Kubernetes and scalable cloud services, FallSafe ensures resilience, efficiency, and future-ready performance for all users.
The FallSafe Device is an IoT-enabled motion tracking system designed for Fall Risk Assessments. It consists of an ESP32-S3 microcontroller, a GY-91 MPU9250 sensor module, and a custom battery pack soldered to a Micro USB connector. This device collects real-time body movement data and transmits it to the cloud for analysis.
- WiFi and Bluetooth-enabled microcontroller with high processing power.
- Facilitates real-time data streaming to the cloud via AWS IoT Core.
- Supports MQTT protocol for efficient message queuing.
- Tracks accelerometer, gyroscope, magnetometer data.
- Provides precise motion tracking for Fall Risk Assessments.
- Made from three AA Energizer batteries in series.
- Soldered to a Micro USB connector for easy ESP32-S3 power supply.
- Enables portability and independent operation.
The FallSafe Device integrates with AWS IoT Core using MQTT, allowing for secure, real-time data transmission. The collected motion data is processed and analyzed in the FallSafe Cloud Application, enabling body motion tracking and fall risk evaluation.
-
Sensor Data Collection
- The GY-91 MPU9250 records movement and positional data.
- Data is processed by the ESP32-S3.
-
MQTT Message Queue
- The ESP32-S3 publishes motion data to an MQTT topic in AWS IoT Core.
- The cloud application subscribes to the motion tracking queue.
-
Data Processing & Dashboard Visualization
- Data is analyzed in real-time.
- Users receive feedback via the FallSafe Dashboard.
- AI-generated insights guide users on improving stability and mobility.
- Lee Guang Le, Jeffrey (Product Owner, Developer)
- Jovan Ong Yi Jie (Developer)
- Putera Daniel (Developer)
- Tan Guo Zhi Kelvin (Developer)