Drexel University

College of Computing and Informatics

Caregiver Indoor Localization

Research
Using BlueTooth to solve the problem of indoor localization for the use case of residential caregiving.
Precise indoor geolocation through BlueTooth has the potential to enable remote or completely digital assistance to those in need of personal caregivers. However, past work regarding the geolocation aspect has identified issues with both accuracy and time delay. This project aims to determine the viability of applying a learning algorithm to this problem, and if applicable, develop this algorithm. If a learning algorithm proves to not be the solution needed, assessments of alternative approaches will be developed. These alternative approaches may involve different hardware, transmission frequencies, and/or another locating algorithm.

Per the recommendation of Dr. Stuart, we focused on Kalman Filtering. Kalman filtering is a linear estimation algorithm that will estimate the state of a dynamic system. In this case, the dynamic system is the device and receiver of care moving around the residence relative to the beacons. Kalman filtering will give us the ability to work around the noise in the RSSI we get from the beacons. Based on the conclusion to use Kalman Filtering, we have begun to implement the algorithm. We were able to leverage what we had already written in Swift for testing the beacons. Building from that, we have the program collecting the necessary information to employ Kalman Filtering. We are able to collect multiple readings from different beacons and store the RSSI and time received in a way that lends itself to Kalman filtering.
...
...

Team Members

...

Behind The Scenes

...
...