Personal Weekly Update | Carolyn | 4/26
Accomplishments
This week, I made good progress integrating Apple Watch functionality into the Cyclify app to collect real-time biometric data during cycling sessions. I encountered roadblocks in doing so, particularly with downloading the app to my watch and establishing a connection between the watch and Xcode simulator. After troubleshooting – looking at Apple’s tech forums, performing software updates, restarting my devices, and reconfiguring my devices, I was finally able to resolve these connectivity issues and can download the Cyclify watch app onto my watch. The watch connection status is displayed in the app interface. My next step is to test this functionality during cycling rides to evaluate how effectively the biometric data transmits from the watch to the app during an actual ride, ensuring the integrated system functions as designed.
On the data visualization side, I have been able to improve the heat map visualization that dynamically displays pressure intensity at the 13 sensor positions, allowing riders to see how their force is being applied during rides. One of the challenges I enountered was the accuracy in the seat representation, it was difficult getting the proportions of the seat to match visually. Another challenge involved correctly positioning the sensors to match their actual physical placement. The visualization now accurately displays pressure variations post rides, allowing cyclists to identify posture patterns.
For the purpose of the demo, I worked on adding visual alerts this week, they appear as a small banner-like notification when the pressure distribution becomes unbalanced. Since the demo will be on a stationary bike, I’ve added immediate visual feedback to show the alerts in case it is loud in the demo area. While this could be more dangerous and distracting during actual road cycling, this implementation is to demonstrate the monitoring capabilities and allows demo participants to see the app’s responsiveness in monitoring the real-time alerts.
Lastly, I researched various voice AI technologies. I liked OpenAI’s text-to-speech AI for fitness applications. After testing its capabilities, I liked the natural and motivating quality of the voice prompts. However OpenAI’s solution requires WiFi, which conflicts with our system requirement of functioning without internet. Apple’s native AVSpeechSynthesizer is compatible offline but offers limited voice options and expressiveness. I’m concerned the AVSpeechSynthesizer’s alerts won’t provide the motivational impact of a fitness instructor. This is a design decision we will debate as we balance connectivity requirements and voice quality.
Progress Status
I’ve made good progress this week and am on track for the final demo. I have to adjust some of the final data visualizations as I finish the Apple Watch integration. But I am excited for the upcoming demo this week and to share all the progress and work we have made!
Next Steps
This week is the final stretch and the last steps include finalizing the text-to-speech voice selection and integration, finishing up the data visualizations and apple watch, as well as flushing out any final design additions and UI additions. A good chunk of this week will also be spent on the final deliverables, final testing, and integration with the team.
