openWearable

A real-time hardware/software platform for wearable robotics development based on the BeagleBone platform.

Overview

openWearable is an open-source project designed to provide a complete hardware and software solution for developing wearable robotic systems, with a focus on real-time control. The platform utilizes the TI BeagleBone Black's Programmable Real-time Units (PRUs) to achieve deterministic high-speed control loops alongside a full Linux environment.

Current applications include:

• Upper limb soft exosuits
• Powered ankle-foot prostheses
• Haptic feedback systems
• EMG/IMU sensing and processing

Getting Started

Hardware Requirements

• BeagleBone Black or BeagleBone Blue
• Compatible sensors (see hardware directory)
• Actuators (PAMs, motors, etc.)

Software Dependencies

• Recommended OS: Debian Linux 12.x (Bookworm) or later
• Kernel: 5.10.168-ti-r72
• PRU Code Generation Tools

Installation

1. Clone the repository:

   git clone https://github.com/jonreal/openWearable.git
   cd openWearable
   

2. Build the library first:

   cd library && make
   

3. Build an application:

   cd ../apps/template && make
   

4. Run the application:

   ./bin/template
   

Device Tree Setup

The device tree configures BeagleBone hardware resources for openWearable. Follow these steps:

1. Clone the BeagleBoard-DeviceTrees repository:

   git clone https://github.com/beagleboard/BeagleBoard-DeviceTrees.git ~/BeagleBoard-DeviceTrees
   

2. Create a symbolic link to your device tree file:

   For BeagleBone Black:

   cd device-tree
   ln -s $PWD/am335x-boneblack-openWearable.dts ~/BeagleBoard-DeviceTrees/src/arm/
   

   For BeagleBone Blue:

   cd device-tree
   ln -s $PWD/am335x-boneblue-openWearable.dts ~/BeagleBoard-DeviceTrees/src/arm/
   

3. Compile and install the device tree:

   cd ~/BeagleBoard-DeviceTrees
   make install_arm
   

4. The compiled device tree files will be installed to /boot/dtbs/<kernel>/ (e.g., /boot/dtbs/5.10.168-ti-r72/)

5. Edit /boot/uEnv.txt to enable the overlay:

   sudo nano /boot/uEnv.txt
   

   Add or modify the following line:

   dtb=am335x-boneblack-openWearable.dtb   # For BeagleBone Black
   # OR
   dtb=am335x-boneblue-openWearable.dtb    # For BeagleBone Blue
   

6. Reboot the BeagleBone:

   sudo reboot
   

Network Setup

To connect to the BeagleBone to Wi-Fi, follow these steps:

iwctl

 [iwctl]# station wlan0 scan
 [iwctl]# station wlan0 get-networks
 [iwctl]# station wlan0 connect <SSID>

Project Structure

• apps/ - Application examples and templates
  • template/ - Starting point for new applications
  • Various application examples (see below)
• library/ - Core libraries
  • include/ - Header files
  • src/ - CPU-side source files
  • pru_src/ - PRU-specific source files
• device-tree/ - Hardware configuration
• hardware/ - PCB designs and hardware documentation
• gui/ - Visualization tools
• utils/ - Utility programs
• scripts/ - Helper scripts

Application Development

Each application consists of these core files:

• state.h - Memory structure definitions
• format.c - Logging format definitions
• cpuloop.c - Application CPU logic
• uiloop.c - User interface logic
• pruloop.c - PRU-specific logic

To create a new application, copy the template directory:

cd apps
cp -r template my_new_app
cd my_new_app

Build Commands

• First build the library: cd library && make
• Then build the app: cd apps/[app-name] && make

Recent Enhancements

1. Build System Reorganization (March 2025)

   • Implemented libowpru.a for sharing PRU driver code
   • Each app links against this shared library

2. DMA for Data Transfer (March 2025)

   • Added DMA support for logging with performance monitoring
   • Use -d flag to disable DMA for benchmarking

3. Binary Logging (In Progress)

   • Converting text-based logs to binary format
   • Improving performance by reducing formatting overhead

Example Applications

• pamtest: Test pneumatic artificial muscles (PAMs)
• haptic-tracking: Haptic feedback control system
• buttontest: Simple I/O demonstration
• imutest: Test IMU sensors and data processing
• reflex: Demonstrate reflex control algorithms

Publications Using openWearable

Soft Wearable Robots for Pediatric Movement Disorders

• Realmuto J, Sanger TD. "Assisting Forearm Function in Children With Movement Disorders via A Soft Wearable Robot With Equilibrium-Point Control." Frontiers in Robotics and AI, 2022; 9:877041. doi:10.3389/frobt.2022.877041.
• Repository branch: equilibrium-point-study

Powered Ankle-Foot Prostheses

• Realmuto, J. "Towards Personalized Powered Ankle-Foot Prostheses." Dissertation, 2018.
• Repository branch: poweredAnkle

Haptic Feedback for Lower Limb Prostheses

• Sie A, Realmuto J, Rombokas E. "A Lower Limb Prosthesis Haptic Feedback System for Stair Descent." ASME Design of Medical Devices Conference, 2017. doi:10.1115/DMD2017-3409.
• Repository branch: hapticFeedback