ElectroGrasp: Electrotactile Aids for Visually Impaired Individuals in Anticipatory Planning and Control of Grasp

Paper Title

ElectroGrasp: Electrotactile Aids for Visually Impaired Individuals in Anticipatory Planning and Control of Grasp

Publication Info

• Topic area: Assistive haptic technology for visually impaired individuals
• Keywords: Electrotactile feedback, visually impaired, anticipatory grasp, tactile perception, proprioception, wearable device, object recognition, grasp planning, haptic assistive technology, motor control

Background and Problem

• Problem / challenge: Visually impaired (VI) individuals lack access to visual cues critical for anticipatory grasp planning, leading to excessive hand aperture, regrasping, and inefficient object interaction. Existing haptic devices primarily address object localization rather than guiding grasp execution.
• Significance: Enhancing grasp planning for VI users can improve their confidence, efficiency, and independence in daily tasks, reducing reliance on compensatory strategies.
• Motivation and related work: Prior research has explored tactile and proprioceptive mechanisms for object perception and developed haptic devices for navigation and object detection. However, these systems often fail to provide functional grasp-affordance cues directly to the hand. This paper addresses this gap by focusing on tactile feedback for anticipatory grasp planning.

Solution

• Proposed approach: ElectroGrasp, a wearable electrotactile system, delivers anticipatory tactile and proprioceptive cues encoding object orientation, size, and shape to support grasp planning and execution.
• Novelty:
  1. Introduction of tactile feedback during the pregrasp phase to compensate for the absence of vision.
  2. Development of dynamic electrotactile encoding for object geometry, size, and orientation.
  3. Integration of proprioceptive and cutaneous cues for anticipatory motor planning.
  4. Demonstration of reduced overshoot and regrasp behaviors, approximating visually guided grasping.
• Procedure and key techniques:
  • Hardware: Flexible electrode arrays on the fingertips and palm deliver high-resolution electrotactile stimulation.
  • Encoding schemes:
    • Orientation: Horizontal/vertical cues via finger and palm stimulation patterns.
    • Size: Modulated stimulation frequency based on finger flex degree.
    • Shape: Data-driven tactile patterns derived from pressure distribution datasets.
  • Evaluation: Three experiments assessed pattern recognition, spatial perception, and grasp execution under electrotactile and audio cueing.

Results

• Concrete findings:
  • Pattern recognition accuracy ranged from 82% to 100%.
  • After 10.5 minutes of training, object recognition accuracy reached 87%.
  • Electrotactile feedback reduced regrasp events from 1.93 (audio) to 1.61 and eliminated overshoot events.
• Advantage over baselines:
  • Electrotactile cueing enabled more accurate anticipatory grasp planning and reduced compensatory movements compared to audio cueing.
  • Provided embodied, spatially mapped cues that directly informed hand posture and grasp configuration.
• Experiments / evaluation:
  • Experiment 1: Assessed pattern discriminability, size sensitivity, and orientation encoding preferences (16 VI participants).
  • Experiment 2: Evaluated object recognition accuracy with and without training (12 VI participants, 18 objects).
  • Experiment 3: Compared grasp execution under audio and electrotactile cueing (15 VI participants, 10 objects).
  • Metrics: Recognition accuracy, regrasp/overshoot events, task duration, NASA-TLX workload scores.
• Limitations and future work:
  • Limited participant pool and short-term evaluations.
  • Wired connections and reliance on external tracking systems reduce wearability.
  • Current encoding schemes lack support for complex shapes, textures, and multi-orientation objects.
  • Future directions include personalized cueing, richer tactile patterns, and fully wearable, self-contained systems.

Summary

ElectroGrasp is a wearable electrotactile system designed to support visually impaired individuals in anticipatory grasp planning by delivering tactile cues for object orientation, size, and shape. The system demonstrated high recognition accuracy, reduced regrasp and overshoot events, and improved spatial perception with minimal training. While current limitations include wired connections and limited encoding complexity, the findings highlight ElectroGrasp's potential as a foundation for future haptic assistive technologies, offering enhanced accessibility and independence for VI users.