L24 Rehabilitation Robotics Flashcards
Rehabilitation Robotics
- Intelligent Mobility Aides
- Robot Manipulation Aides
- Therapeutic Robots: Upper and Lower limb
Intelligent Mobility Aids
- Smart power wheelchairs
- Smart power assist module (SPAM)
- Smart wheeled walkers: Guido
- Sensors: sonar, infrared range finder, laser rangefinder
- Control software
Hands-Free Wheelchair Prototype I
- School of Theater and Dance, CoTA
- Designed for a mixed ability dance group
- Device provides a freedom of movement for a variety of choreographic elements
- First prototype: converted the chair into a large joystick, where leaning to any direction created a motion towards that direction
Hands-Free Wheelchair Prototype II
Second prototype:
* used on almost any commercial powered wheelchair
* Provides wireless control of the powered wheelchair using Bluetooth
* Custom control software for Android based phones
* Accelerometer in phone determine tilt
* Converted to speed power WC controller
Hands-Free Wheelchair - Future
- ultrasonic sensors mounted around the wheelchair
- ultrasonic sensors: conical sensors, each sensor giving general positional data
- an input data source for feedback
- detect obstacles to be avoided
Robotic Manipulation Aids
- Task specific devices: powered feeders, page turners
- Workstation-based manipulation aides: robotics manipulator built into a workstation
- Wheelchair mounted manipulation aides: at USF WMRA (Wheelchair Mounted Robotic Arm)
Workstations and Assistive Feeders
DeVAR Workstation: Stanford
* Excel at one task or task set
* Confines user independence to one
location
MySpoon Feeder: Secom
* Must be installed or setup by caregiver
* User cannot relocate the device
Rail-Mounted Robotic Arms
- Large hardware installation
- Confines user to one location
Mobile Robots and Manipulators
- Interaction at a distance
- Free to move about the environment
- Safe manipulators
- Control and automation in development
Wheelchair-Mounted Robotic Arms
WMRA-II: Univ. of S. Florida
* Reference frame is natural
* Mounting location convenient for
personal tasks
Weston: Bath Institute
* Applicable for general ADLs
* Safe manipulators
* Social impact is critical
WMRA Designs
Two prototype WMRAs have been developed:
* Both use custom 7-DoF robotic arm and 2-DoF power wheelchair for
combined 9-DoF system
* Outperform WMRAs using commercially-available arms [2]
* 7-DoF arm provides capability to overcome singularities, joint limits, and
workspace limitations while performing ADL tasks
iARM by Exact Dynamics
- Product of Exact Dynamics of the Netherlands
- 6-degree of freedom WMRA with 2-finger end effector
- Weight: 9 kg
- Input devices range from 16-button keypads of varying dimension, to 2D joystick, to single-button control.
- Features highly customizable control scheme in which the end user may store up to 12 arbitrary end effector positions and return to these positions with push-button control
JACO by Kinova
*Product of Kinova of Canada
*6-degree of freedom WMRA with 3-finger end effector
*Weight: 5 kg
*Features a weatherproof design and claims to consume less
energy than a standard light bulb
*Standard controller is a 3-axis, or 3D joystick
*Features a high level of singularity and collision avoidance
WMRA Gripper Design
- 1 DOF Parallel motion mechanism.
- Variable 0 to 10 lb force.
- Can handle spherical and lever type doorknobs.
- Can handle switches and push buttons.
- Can handle round and square objects.
- Can handle thin and tiny objects.
- Compatible with the WMRA control system
Vision-Based System for Selecting Objects Using a Brain-Computer Interface (BCI)
- A Brain Computer Interface (BCI) for communicating the needs of a person with severe disabilities (e.g. ALS) to a computer using the brain’s electrical activity
- Using the oddball paradigm with P300 Event Related Potentials (ERPs)
- Flashing objects stimulate
the user’s EEG signal. - User sees likely stimuli
(i.e. the object the user
wants) as well as unlikely
stimuli (i.e. irrelevant
objects or blank space) - A custom BCI System
- Perfroms semi-autonomously activities of daily living (ADLs)
- Minimum required interactions with the user
- Maximize accuracy by custom tailoring the view of the environment shown
- User selects a desired object and task robot performs the task autonomously
Vision-Based System for Selecting Objects Using a Brain-Computer
Interface (BCI) - Modes
Move Mode:
* Select the general area of the object of interest.
* The robotic platform will try to navigate there automatically
Selection Mode:
* User selects the cell whose blue dot
lies on the object
- Using SIFT (Scale Invariant Feature Transform), the system will attempt to
identify the object of interest and display appropriate task choices - For instance, a light switch may generate options to turn the light on or off.
BaxBot
- Baxter robot mounted on PowerBot mobile platform.
- Navigation through unstructured environment and mapping/localization.
- Interactive demos include: follow me; play connect-four; facial recognition; facial expression; get information off the internet; Assist in activities of daily living
BaxBot Teleoperation with Haptic Devices
User Interface
* Three cameras are on BaxBot’s head and both end-effectors, and view angles of the cameras are controllable by the user
* Two Phantom Omni devices provide haptics experience
* IR-sensor graph provides the distance information between objects and end- effectors.
* Velocity control based on two modes:
- Base frame control
- End-effector frame control
Autonomous Pick and Play
- Use 3D computer vision to recognize
objects based on their geometries and estimate their 6D poses - Use a database of object 3D models and default grasp options
- Tablet touch screen user interface
Human-Robot iPhone Interface
- Regular joysticks cannot provide 6 DoF control.
- User friendly interface is designed using the iPhone’s accelerometer, gyroscope and touch screen.
Lower limb rehabilitation
- Hocoma Lokomat
- Robotic gait orthosis
- Used to treat individuals with neurological diseases or injuries such as stroke, SCI, brain injury, MS and cerebral palsy
Upper Limb Rehabilitation
Forced use
* Effectively retrains highly functioning people with stroke
Robotic Rehabilitation
* Many have been developed - a couple examples are: MIT-MANUS, T-WREX
Two general methods of robot-rehabilitation:
* Assistive - guide the user: trains coordination
* Resistive - push against force: trains strength
Bimanual Upper Limb Rehabilitation
- Tight neural coupling allows the same
motor commands to be duplicated on the opposite side of the body. - Physical coupling provides same motions and feedback for both limbs.
