Navicate: Development of a Smart Cane with Time-of-Flight Sensors for Elevation Change Detection, and Machine Learning-Based Object Recognition

Document Types

Paper Presentation

Research Theme (for Paper Presentation and Poster Presentation submissions only)

Computer and Software Technology, and Robotics (CSR)

School Name

Mindanao State University - General Santos City, College of Education Training Department, Senior High School Department

Track or Strand

Science, Technology, Engineering, and Mathematics (STEM)

Research Advisor (Last Name, First Name, Middle Initial)

Crusis, Algem Cris, B.

Start Date

25-6-2026 10:30 AM

End Date

25-6-2026 12:00 PM

Zoom Link/ Room Assignment

Online - https://zoom.us/j/91936856247?pwd=oCMfMsh44I2wb0dYsEgoInDJy59bOq.1 Meeting ID: 919 3685 6247 | Passcode: research

Abstract/Executive Summary

Mobility and environmental awareness remain significant challenges for visually impaired individuals, particularly in detecting elevation changes and identifying potential hazards. This study presents Navicate, a smart cane developed to improve safe navigation through the integration of dual Time-of-Flight (ToF) sensors and machine learning–based object recognition using TensorFlow Lite. The device utilizes a Raspberry Pi 4 to process depth and visual data, providing real-time haptic and audio feedback to inform users of ground-level changes and recognized objects. Navicate was tested in controlled indoor settings to evaluate its performance. The ToF sensors were tested on stairs and inclined surfaces to determine their capability to detect depth variations, while the machine learning model was evaluated using selected object categories relevant to everyday navigation. Data were analyzed to compute average response times and success rates. Results show that both ToF sensors achieved a 100% accuracy rate in detecting elevation changes. The vibration motors responded within an average of 27.78 (left) and 31.11 (right) milliseconds, suggesting very good responsiveness. Audio feedback showed similar response times, while the machine learning model produced 100% classification accuracy, demonstrating high reliability in real-world navigation. These findings suggest that Navicate has strong potential for enhancing user safety, reducing navigation risks, and promoting independence. Overall, the study contributes to the advancement of assistive technology by offering a scalable, sensor-integrated framework that may guide future research and innovation in smart mobility systems.

Keywords

assistive technology; Time-of-Flight sensors; machine learning; object recognition; smart cane navigation

Statement of Originality

yes

This document is currently not available here.

Share

COinS
 
Jun 25th, 10:30 AM Jun 25th, 12:00 PM

Navicate: Development of a Smart Cane with Time-of-Flight Sensors for Elevation Change Detection, and Machine Learning-Based Object Recognition

Mobility and environmental awareness remain significant challenges for visually impaired individuals, particularly in detecting elevation changes and identifying potential hazards. This study presents Navicate, a smart cane developed to improve safe navigation through the integration of dual Time-of-Flight (ToF) sensors and machine learning–based object recognition using TensorFlow Lite. The device utilizes a Raspberry Pi 4 to process depth and visual data, providing real-time haptic and audio feedback to inform users of ground-level changes and recognized objects. Navicate was tested in controlled indoor settings to evaluate its performance. The ToF sensors were tested on stairs and inclined surfaces to determine their capability to detect depth variations, while the machine learning model was evaluated using selected object categories relevant to everyday navigation. Data were analyzed to compute average response times and success rates. Results show that both ToF sensors achieved a 100% accuracy rate in detecting elevation changes. The vibration motors responded within an average of 27.78 (left) and 31.11 (right) milliseconds, suggesting very good responsiveness. Audio feedback showed similar response times, while the machine learning model produced 100% classification accuracy, demonstrating high reliability in real-world navigation. These findings suggest that Navicate has strong potential for enhancing user safety, reducing navigation risks, and promoting independence. Overall, the study contributes to the advancement of assistive technology by offering a scalable, sensor-integrated framework that may guide future research and innovation in smart mobility systems.

https://animorepository.dlsu.edu.ph/conf_shsrescon/2026/BoA_CSR/5