Personal Accomplishment
What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week (12+ hours).
On the software side, I finished a working prototype of the website with React. This included connecting Google Authentication and the Google Maps API to manage user login and real-time location tracking. Google Authentication is our choice of login since it is secure and easy to use. We choose React for the front end for its modularity, allowing us to maintain clear code structure and scalability. Integrating the Google Maps API also gives us reliable, real-time location tracking incase our gps component end up failing. Our GPS is performing well outdoors but currently fails indoors due to typical limitations of GPS with line-of-sight requirements.
Image for web app prototype:
Hardware progress this week included testing and initial calibration of both the radar (KLD7) and the accelerometer. The radar testing revealed angular limitations aligning with the specifications in the data sheet, which confirms that detection reliability drops off at ±80° horizontally. During testing, we observed that vertical detection was limited to a 30° coverage range, as specified in the data sheet. To better understand these boundaries, we conducted both indoor and outdoor tests across various distances and angles. Indoor tests showed limitations in obstacle detection, as the radar struggles to detect motion when obstructed by physical barriers (e.g., tables, computers, chairs), which is expected for this sensor type. The outdoor testing was divided into defined sectors (left, mid, right) and distances (1m, 3m, and 5m) Notably, the radar consistently picked up signals within the set parameters, but some sections (e.g., -60°, 3m, and -20°, 4.5m) showed intermittent detection issues that require further exploration, likely due to signal interference or sensitivity at certain angles.
For the accelerometer, we setup the code in C, which allowed detection of initial movement data, indicating that the device is wired and communicating as expected. However, raw movement detection isn’t sufficient for the intended fall detection functionality, which requires more nuanced data handling to differentiate between minor movement and actual falls. Additional tuning is necessary to filter and threshold accelerometer signals to isolate patterns unique to falls. I anticipate that refining these settings will involve adjusting sensitivity thresholds and possibly implementing an averaging algorithm to reduce noise from incidental movements.
In preparation for the final demo, I ordered a white cane, which will help simulate real-world use cases which is emphasizing the practical safety applications of WalkGuard for visually impaired users.
Schedule
Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
I am currently on schedule with my project. The website prototype is functioning as intended, and the hardware setup is progressing. I anticipate completing the necessary tweaks to the accelerometer in the coming days.
Upcoming Deliverables
What deliverables do you hope to complete in the next week?
I’ll focus on tuning sensitivity to detect falls accurately, adjusting settings to distinguish significant movement patterns from minor fluctuations. Also I aim to enhance our website’s functions and enable connection between our website and the rpi using api, considering both user’s phone and the device is using the same network. Finally I plan to map specific distance markers on the floor to standardize our testing parameters, allowing us to identify and potentially resolve inconsistencies in detection at certain angles or distances.