Personal Weekly Update | Zoe | 4/12

Accomplishments

This week I worked on expanding and improving both the sensing and imaging systems. I implemented 16 total force sensors using multiplexers and got them working through the ESP32. These will be mounted on the bike tomorrow so we can begin calibrating with the full setup. I’ll be flashing the ESP32 and passing it off to Carolyn to make sure the app still functions correctly with the upgraded sensor array during both calibration and ride modes.

The calibration process itself also needed a few changes. I updated the logic to include an additional posture pose to help capture proper riding form. We also added physical weights during the calibration phase to improve the quality of the sensor data and better separate the poses.

On the imaging side, I set up and tested both the Raspberry Pi camera and a standard webcam. Both are working, but additional testing is needed to compare frame quality, latency, and mounting feasibility. We’ll also need to evaluate whether each camera works better from the top or bottom of the new printed housing that Rita redesigned this week.

Progress Status

We’re slightly behind where I wanted to be this week in terms of final mounting, but the key functionality is working. The 16-sensor system is live, and the imaging setup is operational. We’ll be able to catch up once the new sensor layout is installed on the bike, and we begin full-system calibration and app-side testing.

Next Steps

– Mount the 16 sensors and multiplexers to the bike
– Begin full calibration using the updated posture logic
– Pass flashed ESP32 to Carolyn for app testing and BLE compatibility
– Test webcam vs. Pi camera performance under actual riding conditions
– Finalize decision on camera placement with the printed housing

Verification Plan

For verification, I will focus on confirming that the force sensor subsystem meets our engineering and use case requirements. This includes:

– Verifying that all 16 sensors provide stable, unique values during calibration poses
– Confirming that sensor data can be collected consistently via the ESP32 without BLE packet loss or timing issues
– Running calibration sessions to verify that each pose produces distinguishable sensor patterns, especially with the added weight-based separation
– Testing the BLE interface to confirm that real-time ride data and calibration storage still work with the larger sensor array

We’ll compare calibration data logs to ensure that each sensor channel is being used and that the values match expected thresholds under known loads. For imaging, we will run side-by-side trials of the Pi camera and webcam under the same lighting and movement conditions, using captured frame samples and timestamps to analyze latency and clarity. The final goal is to confirm that the full system, including sensing and imaging, meets both the responsiveness and stability needed for the intended rider use case.