Smart Walker

A mobile/desktop app + embedded hardware project for the Spring 2026 UW Madison BME Design Smart Walker project. The Flutter app targets iOS, Android, macOS, and Windows x64 and connects via BLE to three sensor modules mounted on a walker — two ESP32 load cells (feet) and a Raspberry Pi + TI AWR1843 mmWave radar — to display live weight, distance, and walking speed data. Trials can be recorded per patient and exported as CSV.

Features

Live BLE Telemetry

• Live left, right, and total weight (lbs), distance (ft), and speed (ft/s).
• Automatic discovery and reconnection for the three BLE sensor modules.
• Compact sensor-readiness indicators and retry controls.

Trial Recording

• Persistent bottom start/stop control and an elapsed timer (MM:SS.cc) while recording.
• Samples captured at a nominal 200 ms interval.
• Live walking-speed and left/right weight charts showing the latest 30 seconds; the saved trial retains the complete recording.
• Trials stop and preserve partial data when required telemetry is lost or the app enters the background. Recording is foreground-only.

Patient Management

• Add and select patients, or record under the default Unassigned patient.
• Trials are attached to the patient selected when recording starts.
• Patient and trial deletion require confirmation.

Trial History

• Patient-specific history with duration, distance, average speed, and average weight.
• Full-trial speed and weight-distribution charts.
• CSV file export through the platform share interface.
• Local JSON persistence, previous-save backups, and interrupted-recording recovery.

Project Structure

Smart Walker/
├── AGENTS.md                   # Repository-wide agent, setup, and deployment guidance
├── app/                        # Flutter + Forui app (iOS / Android / macOS / Windows)
│   ├── lib/                    # UI, models, app services, BLE and persistence
│   ├── android/                # Android permissions and native host
│   ├── ios/                    # iOS permissions, signing, and native host
│   ├── macos/                  # Desktop window, Bluetooth entitlements, native host
│   ├── windows/                # x64 host, native Save As, storage/instance protection
│   └── pubspec.yaml            # App version and dependencies
└── controller/
    ├── swLeftFoot.ino          # ESP32 left foot load cell → BLE
    ├── swRightFoot.ino         # ESP32 right foot load cell → BLE
    ├── swRadar.py              # Raspberry Pi AWR1843 radar via USB → BLE
    └── AWR1843config.cfg       # TI mmWave radar configuration

Acknowledgments

• Radar data parsing based on AWR1843-Read-Data-Python by Ibai Gorordo (MIT License)