SS-Project: Hall Sensor Research

A research project for Hall Sensor developed by Jaehong Cho in 2025. This project includes sensor data acquisition, calibration, training, and real-time force visualization.

Installation

Step 1: Setup Conda Environment

Create a conda environment using the provided environment.yml file:

conda env create -f environment.yml
conda activate SS
export PYTHONPATH=$(pwd):$PYTHONPATH

Step 2: Install PyTorch

Install PyTorch based on your system configuration. Visit PyTorch official installation guide for the latest instructions.

Example command for PyTorch 2.6.0 & CUDA 11.8:

pip install torch==2.6.0 torchvision==0.21.0 torchaudio==2.6.0 --index-url https://download.pytorch.org/whl/cu118

For CPU-only:

pip install torch==2.6.0 torchvision==0.21.0 torchaudio==2.6.0

Note: Choose the appropriate CUDA version for your system or use CPU-only installation.

Hardware Setup

Step 1: Connect Sensors

• Hall Sensor (sensor_srbl): Connect to /dev/ttyUSB0
• Commercial FT Sensor (RFT60-HA01): Connect to /dev/ttyUSB1 (optional, only required for calibration)

Step 2: Verify USB Connections

Check USB device connections using:

dmesg | grep tty

Usage

Step 1: Test Hall Sensor

Test the Hall sensor to ensure proper connection and functionality:

python sensor/sensor_srbl.py

Step 2: Test Commercial FT Sensor (Optional)

If you have a commercial FT sensor connected, test it:

python sensor/sensor_FT.py

Step 3: Collect Calibration Data

Collect sensor data for calibration training:

python calibration/logger.py

This will generate a CSV file with synchronized data from both sensors.

Step 4: Train Calibration Model

Train the neural network model for force prediction:

python calibration/train.py

This will create trained model weights and scalers in the calibration/ directory.

Step 5: Visualize Sensor with Force Mapping

Run the real-time force visualization using VTK:

python run_vtk.py

You can optionally specify an STL file for the sensor model:

python run_vtk.py sensor.STL

Project Structure

SS-Project/
├── sensor/                    # Sensor interface modules
│   ├── sensor_srbl.py        # Hall sensor driver
│   └── sensor_FT.py          # Commercial FT sensor driver
├── calibration/              # Calibration and training scripts
│   ├── logger.py             # Data logging script
│   ├── train.py              # Model training script
│   └── sensor_data_log_*.csv # Generated calibration data
├── run_vtk.py                # Real-time force visualization
└── environment.yml           # Conda environment file

License

Research Project - Soft Robotics & Bionics Lab, Seoul National University