Contactless, real-time health monitoring powered by iPhone camera and AI voice guidance.
Built in 6 hours at the CrimsonCode Hackathon 2026 at Washington State University.
HealthCam uses an iPhone's front-facing camera to contactlessly measure pulse rate, breathing rate, and stress levels via remote photoplethysmography (rPPG). An AI agent powered by ElevenLabs delivers live verbal de-escalation guidance in real time. Every session is automatically analyzed and logged.
No wearables. No extra hardware. No friction.
Health problems can escalate fast. HealthCam was built to give officers and individuals a real-time edge — physiological data on a subject's stress state — before things go wrong.
- Face Detection — The iPhone camera captures the user's face using AVFoundation and CoreImage.
- Biometric Extraction — The Presage SmartSpectra SDK performs rPPG analysis to extract pulse, breathing rate, and stress indicators from subtle skin-color changes.
- AI Voice Agent — ElevenLabs Conversational AI SDK processes the biometric data and delivers real-time, voice-first de-escalation guidance.
- Session Logging — Every interaction is logged and analyzed for post-session review.
| Layer | Technology |
|---|---|
| Language | Swift, SwiftUI |
| IDE | Xcode |
| Biometrics Engine | Presage SmartSpectra SDK (rPPG) |
| Voice AI | ElevenLabs Conversational AI SDK |
| Real-Time Communication | LiveKit |
| Frameworks | AVFoundation, CoreImage, Combine |
| Serialization | SwiftProtobuf |
crimson-code-26/
├── SmartSpectra-swift-sdk/ # Presage SmartSpectra SDK integration for rPPG biometric detection
├── core/ # Core iOS application — SwiftUI views, AI agent, session logic
- Getting Presage's headless processing pipeline working without their native UI — the SDK's recording gate only opens when
statusCode == .ok. - ElevenLabs session persistence required separating intentional disconnects from post-turn state changes.
- rPPG is extremely sensitive to lighting conditions and face positioning.
- 📹 Body cam integration
- 👥 Multi-subject tracking
- 📊 Department-level analytics dashboard
- 🧠 PTSD early-warning indicators for officers
- macOS with Xcode installed
- iOS device with a front-facing camera (simulator not supported for camera-based rPPG)
- Presage SmartSpectra SDK credentials
- ElevenLabs API key
-
Clone the repository:
git clone https://github.com/1paramveer/crimson-code-26.git cd crimson-code-26 -
Open the project in Xcode:
open core/*.xcodeproj -
Configure your API keys for SmartSpectra and ElevenLabs in the appropriate configuration files.
-
Build and run on a physical iOS device.
HealthCam was developed as part of a collaborative team during the CrimsonCode 2026 Hackathon at Washington State University.
Team Members
- Paramveer
- Aman Verma
- Srishanth
The project was built under typical hackathon conditions where multiple components were developed and integrated rapidly. Team members worked across shared systems and development environments while assembling the final prototype and preparing the demonstration.
For a detailed reflection on the hackathon experience — including technical decisions, contributions, and lessons learned — see the Reflection Document. You can also view the project submission on Devpost.
Built with ❤️ at CrimsonCode 2026 — Washington State University