Transforming Care with Smart Home Technology: Data-Driven Solutions for Group Homes
Written by Tobias Weinberg
Understanding the Problem
Many group homes supporting individuals with disabilities use smart home technology to enhance care and ensure safety. Devices such as motion sensors, cameras, and smart speakers can monitor activity, prevent accidents, and assist staff with daily routines. However, despite the potential benefits, these devices are often deployed in a fragmented way. Data from different devices is typically siloed in separate systems, making it difficult to analyze patterns or measure the overall impact of technology on residents' quality of life.
As a Siegel PiTech PhD Impact Fellow with YAI, I explored how to centralize data collection from various smart devices in group homes. My goal was to create a unified system that could collect, visualize, and analyze the data in real-time, helping staff make better-informed decisions about care and safety. Additionally, such a system could be used to explore how technology influences staffing patterns and overall efficiency in group homes.
Discovery and Exploration: Unifying Smart Devices for Improved Care
Image: Shutterstock
During my fellowship, I collaborated closely with YAI’s IT department and visited several group homes to observe how smart devices were being used in practice. The first step was cataloging the various types of technology in use, such as motion sensors, security cameras, environmental monitors, and smart assistants. Each of these devices served a specific purpose, but their isolated data collection limited their potential impact.
Once I had a clear understanding of the devices in use, I began exploring how to access the data they generated. This required working closely with technology vendors to understand the technical specifications and data formats of each device. One of the challenges was that different vendors often use proprietary data formats, making it difficult to create a standardized approach to data collection. Some devices were more cooperative than others, offering straightforward access to their data, while others required custom integrations or workarounds.
As I delved deeper, I realized that the key to a successful solution would be to create a flexible system that could accommodate data from a wide variety of devices, regardless of vendor or format. This discovery phase highlighted the complexity of the problem but also underscored the importance of building a robust and adaptable data infrastructure.
Choosing the Right Tools to Address the Problem
To create a robust solution to collect data from the various devices used in group homes, I turned to an open-source home automation platform called Home Assistant, which emphasizes local control and privacy, and is supported by a vibrant community of DIY enthusiasts and tinkerers from around the world.
By leveraging Home Assistant, I integrated devices from multiple vendors, including RING (a smart home security company) and Matter (a standard for smart home devices). This enabled me to aggregate data from a variety of sensors into one central system. To make this data usable, I created an API that external websites could request to retrieve and plot the data.
With these tools in place, I built a website that could efficiently handle dynamic data filtering and visualization. Using FLASK and JavaScript, I developed a concept that enables staff members to seamlessly filter data, select specific time ranges, and view real-time information from multiple sources. The prototype provides a more intuitive and interactive experience for managing and analyzing device data.
Example plot with hourly granularity showing decreased access to the lab during weekends. Please note that the data in the plot is lab synthesized data for exemplary purpose.
Impact and Path Forward: Expanding Smart Technology Across Group Homes
Tobias Weinberg
Ph.D. student, computer science, Cornell Tech
The system developed during this fellowship is a significant step towards harnessing the power of smart home technology in group homes. By centralizing data collection and visualization, the platform gives YAI easier access to actionable insights, which can enhance residents' quality of life and improve staff efficiency. This data-driven approach is essential for understanding the broader impact of technology on care and staffing patterns.
Thanks to the generosity of the Siegel PiTech PhD Impact Fellowship, my involvement in this project will extend beyond the summer fellowship. In the next phase, we will deploy the system in a real group home environment and begin analyzing the data collected from the sensors. The goal is to identify behavior patterns from the raw data, which will provide valuable insights into residents' daily routines and how staff can better support them.
As we gather more data, we plan to extend the system to multiple group homes, allowing us to compare patterns across different environments. This will help us determine whether the insights gained from one home can be applied more broadly, potentially leading to organization-wide improvements in care and efficiency.
YAI plans use the data collected to assess the effectiveness of the sensors, understand their impact on behavior and treatment, and ensure the safety of those living in the homes. This ongoing work will also help YAI stay at the forefront of innovation in care for people with disabilities, using technology to create safer, more supportive living environments.