SafeSpot

Main Screen 1 - showing live streams and displaying info about the patients
Main Screen 2 - displaying the alert to attract the nurse's attention
Details 1 - viewing the room and info
Details 2 - viewing the alert with an option of notifying the emergency contact

Problem Statement

As the number of the elderly population is constantly growing, there is an increasing demand for home care. In fact, the market for safety and security solutions in the healthcare sector is estimated to reach $40.1 billion by 2025. The elderly, disabled, and vulnerable people face a constant risk of falls and other accidents, especially in environments like hospitals, nursing homes, and home care environments, where they require constant supervision. However, traditional monitoring methods, such as human caregivers or surveillance cameras, are often not enough to provide prompt and effective responses in emergency situations. This potentially has serious consequences, including injury, prolonged recovery, and increased healthcare costs.

Solution

The proposed app aims to address this problem by providing real-time monitoring and alert system, using a camera and cloud-based machine learning algorithms to detect any signs of injury or danger, and immediately notify designated emergency contacts, such as healthcare professionals, with information about the user's condition and collected personal data. We believe that the app has the potential to revolutionize the way vulnerable individuals are monitored and protected, by providing a safer and more secure environment in designated institutions.

Developing Process

Prior to development, our designer used Figma to create a prototype which was used as a reference point when the developers were building the platform in HTML, CSS, and ReactJs. For the cloud-based machine learning algorithms, we used Computer Vision, Open CV, Numpy, and Flask to train the model on a dataset of various poses and movements and to detect any signs of injury or danger in real time. Because of limited resources, we decided to use our phones as an analogue to cameras to do the live streams for the real-time monitoring.

Impact

Improved safety: The real-time monitoring and alert system provided by the app helps to reduce the risk of falls and other accidents, keeping vulnerable individuals safer and reducing the likelihood of serious injury.
Faster response time: The app triggers an alert and sends notifications to designated emergency contacts in case of any danger or injury, which allows for a faster response time and more effective response.
Increased efficiency: Using cloud-based machine learning algorithms and computer vision techniques allow the app to analyze the user's movements and detect any signs of danger without constant human supervision.
Better patient care: In a hospital setting, the app could be used to monitor patients and alert nurses if they are in danger of falling or if their vital signs indicate that they need medical attention. This could lead to improved patient care, reduced medical costs, and faster recovery times.
Peace of mind for families and caregivers: The app provides families and caregivers with peace of mind, knowing that their loved ones are being monitored and protected and that they will be immediately notified in case of any danger or emergency.

Challenges

One of the biggest challenges have been integrating all the different technologies, such as live streaming and machine learning algorithms, and making sure they worked together seamlessly.

Successes

The project was a collaborative effort between a designer and developers, which highlights the importance of cross-functional teams in delivering complex technical solutions. Overall, the project was a success and resulted in a cutting-edge solution that can help protect vulnerable individuals.

Things Learnt

Importance of cross-functional teams: As there were different specialists working on the project, it helped us understand the value of cross-functional teams in addressing complex challenges and delivering successful results.
Integrating different technologies: Our team learned the challenges and importance of integrating different technologies to deliver a seamless and effective solution.
Machine learning for health applications: After doing the research and completing the project, our team learned about the potential and challenges of using machine learning in the healthcare industry, and the steps required to build and deploy a successful machine learning model.

Future Plans for SafeSpot

First of all, the usage of the app could be extended to other settings, such as elderly care facilities, schools, kindergartens, or emergency rooms to provide a safer and more secure environment for vulnerable individuals.
Apart from the web, the platform could also be implemented as a mobile app. In this case scenario, the alert would pop up privately on the user’s phone and notify only people who are given access to it.
The app could also be integrated with wearable devices, such as fitness trackers, which could provide additional data and context to help determine if the user is in danger or has been injured.