Menu
Home
Log in / Register
 
Home arrow Computer Science arrow Robotic Assistive Technologies: Principles and Practice
Source

Input Methods

Smart wheelchairs have been used to explore a variety of alternatives to the more “traditional” input methods associated with power wheelchairs (e.g., joystick, pneumatic switches). More recent smart wheelchairs have used brain-computer interfaces (Carlson and del R. Millan 2013; Mandel et al. 2009), eye gaze (Rofer, Mandel, and Laue 2009), and voice recognition (Pineau et al. 2011; Simpson and Levine 2002). A summary of alternative input methods can be found in Table 5.2.

Feedback Modalities

Smart wheelchairs can provide feedback to the driver through various modalities and interfaces (Viswanathan, Zambalde et al. 2016; Wang,

Sensor

Capability

Advantages

Disadvantages

Ultrasonic (sonar)

Measures distance to obstacles by emitting sound waves

  • • Small
  • • Low power
  • • Low cost
  • • Sensitive to obstacle properties (such as sound-absorbing material)
  • • Cannot detect very small/thin or concave obstacles
  • • Cross-talk issues with other sounds or multiple ultrasonic sensors
  • • Cannot detect drop-offs unless sensor is pointed to the floor
  • • Limited coverage

Infrared (IR)

Measures distance to obstacles by emitting light; can also detect transitions from light to dark lines

  • • Small
  • • Low power
  • • Low cost
  • • False positives in natural light (due to IR interference)
  • • Sensitive to flooring materials (issues with dark, light-absorbent, transparent, and refractive surfaces)
  • • Limited coverage

Laser range finder

Provides 180° scan of obstacles (distances) in the environment

  • • High precision
  • • Good for longer ranges
  • • High cost
  • • High power
  • • Possible eye safety issues
  • • Only detects objects at the height of the laser range finder

Stereovision

camera

Captures color or black- and-white image and depth images of the environment

  • • Low power
  • • Medium cost
  • • Can be used for high-level scene understanding (e.g., object and location recognition) and head/eye tracking of users
  • • Cannot detect textureless objects
  • • Cannot perform in poorly lit conditions
  • • Challenges posed by reflective and transparent surfaces

(Continued)

TABLE 5.1 (CONTINUED) Summary of Sensors Used by Smart Wheelchairs

Sensor

Capability

Advantages

Disadvantages

Structured light (e.g., Microsoft Kinect)

Captures color or black- and-white image and depth images (using projected IR) of the environment

  • • Low power
  • • Low cost
  • • Can be used for high-level scene understanding (e.g., object and location recognition) and head/eye tracking of users
  • • Challenges posed by outdoor environments (due to IR interference)
  • • Sees through transparent surfaces (i.e., cannot be used to avoid collisions with glass doors)
  • • Can demand a lot of processing power

Bump (tactile sensor)

Can detect if contact is made (force is exerted) by an object

  • • Low cost
  • • Low power
  • • Requires contact (safety issues)
  • • Increasing coverage (e.g., by using bumper skirt-like sensors) can lead to increased form factor and bulkiness

Wheelchair

encoders

Measures number and speed of wheel rotations

• Can be used to easily gather information on distances traveled and speed

  • • Can be expensive to retrofit on wheelchairs
  • • Wheel slippage can lead to errors

Inertial measurement units (IMUs)

Contain accelerometers, gyroscopes, and optional magnetometers that can be used to determine velocity, orientation, and position, respectively

  • • Low cost
  • • Low power
  • • Prone to accumulated errors when used for longer distances
  • • Higher-accuracy IMUs can be expensive

Global positioning system (GPS)

Provides location and time information

  • • Can give accurate location when GPS satellites are in an unobstructed line of sight
  • • Low cost

• Does not work well indoors

TABLE 5.2 Alternative Input Methods Used in Smart Wheelchairs

Input

Capability

Advantages

Disadvantages

Brain-computer

interface

Sensors may be worn on the head (e.g., electroencephalography, EEG) or implanted in the skull (e.g., electrocorticography, ECOG).

• Requires no physical movement

  • • Requires significant computational power
  • • Expensive
  • • Limited signals available for control
  • • Sensors must be placed on or inside the skull

Eye gaze

Cameras pointed at the user’s face track where the user is looking to determine where the PWC should steer.

  • • Requires limited physical movement
  • • No sensors touch the user’s body
  • • Requires significant computational power
  • • Distinguishing between “looking to drive” and “looking to see” is challenging
  • • Expensive

Voice recognition

Speech recognition software interprets steering commands from user.

  • • Low cost
  • • No sensors touch the user’s body
  • • Requires ability to speak clearly and consistently
  • • Voice has limited bandwidth

Mihailidis et al. 2011). For example, speakers can be used to provide audio feedback, while small light-emitting diodes (LEDs) can be used to offer visual feedback (such as a flashing red light if the driver gets too close to an object). Haptic feedback can be provided by using small vibrations to warn the driver about obstacles. Touch screen displays can be used both to gather information from the driver (e.g., desired destination) and to display information (e.g., scheduled activities, way-finding assistance, virtual environments, etc.). Feedback modalities that have been described in the literature are summarized in Table 5.3.

 
Source
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Mathematics
Political science
Philosophy
Psychology
Religion
Sociology
Travel