Team Weekly Update | 4/26
Accomplishments
This week we finished most of the final system testing, finalized our project presentation, and prepared for the upcoming demo. The pressure sensing system, BLE communication, calibration process, and app visualization are all complete and tested.
We completed a full round of testing while the bike was mounted on the stand. From these tests, we validated that pose matching, confidence scoring, and real-time feedback are working as expected.
We also spent time reviewing and editing our final slides, organizing the system explanation from sensing hardware through to app interaction.
The Raspberry Pi camera system setup is still ongoing and will continue to be worked on after the main demo.
Progress Status
The system is complete for the demo and presentation, with 13 sensors installed and functional. BLE communication and calibration logic are working well. Live bike testing shows strong performance while mounted. The imaging component is still being developed and is considered a future extension.
Next Steps
– Deliver final presentation and live demo
– Submit final report and project documentation
– If time allows, test free-riding behavior without the bike stand
– Continue work on the Raspberry Pi camera system
Individual Contributions
Zoe
– Finalized calibration system with sensor masking and intensity scoring
– Completed data analysis and chart creation for presentation
– Tuned BLE stability and helped sync timing with the app
– Finished main presentation slide set
– Worked on soldering, mounting, and debugging the full 13-sensor system
– Continued working on Raspberry Pi camera setup
Rita
– Finalized calibration system testing
– Completed data analysis and chart creation for presentation
– Iterated device housing to accommodate increased pressure sensors
– Finished main presentation slide set
– Determined optimal placing for Raspberry Pi camera
Carolyn
– Integrated Apple Watch functionality
– Troubleshot and resolved watch-to-Xcode connectivity issues, downloaded app on the watch
– Improved the seat heat map visualization for accurate pressure display across 13 sensors
– Added dynamic visual alerts for posture imbalances
– Evaluated and tested voice AI options
Verification and Validation Summary
Unit Tests performed:
– force sensor response testing with tuned resistor values
– BLE packet transfer validation across multiple phones
– calibration matching tests for each pose while the bike was mounted
– intensity matching tests to improve live pose classification accuracy
System Tests performed:
– full calibration and riding tests with 13 sensors installed on the mounted bike
– mounted riding trials to test pose classification stability during pedaling and shifting weight
– app visualization validation during real-time ride sessions
Findings:
– basic sensor masking was not enough for consistent riding pose detection, especially under dynamic shifts
– adding intensity-based scoring helped maintain pose classification even when weight was distributed differently across sensors
– BLE communication was stable across multiple trials and device types
– the system performed well under mounted riding, although free-riding tests are still planned if time allows
Overall, the system reliably detects rider posture in real time, communicates with the app, and gives appropriate feedback based on calibration without requiring major rework. Imaging remains an area for continued development beyond the demo.