This device helps the visually impaired by alerting them of objects nearby and providing directions.

Arnav Mishra

Overview

According to a 2010 WHO report, over 285 million people across the globe are legally blind. Numerous technological solutions are being explored to alert others of the presence of visually impaired people, but there is still a lack of capable assistive technologies to enable the visually impaired to navigate by themselves. Therefore, the goal of this project is to design a solution that allows these individuals to walk outdoors safely and independently through object recognition and path determination. First, the accuracy and precision of the infrared and ultrasonic sensors were tested on aluminum objects placed 50, 75, 100, 200, 300 cm away. These aluminum objects would have eccentricities of infinity (no curvature), 1, and 2 to test how the eccentricity affects the accuracy and precision of the sensors. The eccentricity was tested for both concave and convex curves. Then, an algorithm was tested in a virtual environment with 4, 5, and 6 obstacles randomly placed in a simulated area to represent the global positioning system on a smaller level. Finally, the algorithm was connected to the sensors and tested in the same environment with physical obstacles. The setup for both scenarios was an array of 25 available spaces. For the virtually simulated model, these were measured as nodes and in the physical testing, aluminum objects of 0.5m by 0.5m would be put in an available space of the physical 25 space array. The overall accuracy of both sensors decreased as the eccentricity of the object increased, however, the precision stayed relatively low. A low standard deviation shows the data is accountable and valid conclusions can be made based on it. Comparing the eccentricity of 2 and infinity, the t-value was as much as 11.589 at 300 cm away for the ultrasonic sensor showing a significant change in accuracy farther away. At a closer range, the t-value was as low as 1.481 at 50 cm away for the ultrasonic sensor showing that at a closer distance, there was not a significant change of accuracy dependent on the eccentricity of the object. This suggests that the sensors did not have to be modified to account for the eccentricity as they could still detect objects close by regardless of the curvature. For the obstacle avoidance algorithm, the system was quite accurate overall as the program was successful 94.0% of the time in the computer-simulated environment and 92.0% successful when using the sensors with the A-Star search algorithm in physical object testing. Combined with testing in previous years, the final prototype can detect most materials, colors, and eccentricities. It also has an application with verbal directions based on the algorithm. Some future extensions of this project include human testing for comfort testing of the prototype and adding Google API to navigate with real road maps. This solution bridges a major knowledge gap in the community regarding visually impaired technology and aims to help the visually impaired navigate outdoors more comfortably.

My Story

I have been working on this project for four years now. It all started in my freshman year of high school, and I talked about the rationale behind starting this project. At Shrewsbury High, we had this class called Research Methods that prepares you to make a science fair project and compete at the regional and state fairs. As a freshman, I did not have much expectation to place at such a competitive level, but I advanced from my school all the way to the state competition my freshman year. That year may have been the toughest year I had been through though. In the middle of my research, I had learned that my grandmother had passed away. I was absolutely devastated. I did not know how to continue my project due to the fact that I had to go to India for 3 weeks and I could not really bring my materials there to build my cane. As such, I spent as much time as I could planning what I would do when I returned back to the United States, but I had to cut down the prospects of my project. However, with the time I had, I was glad to have received a 2nd place award both at the regional and state level. The following year, I wanted to continue my research in honor of my grandmother. I worked as hard as I could to create a more developed prototype of my project only to find that COVID shut down schools. I was given the opportunity to compete at the state level but was not given the opportunity to compete for a chance to go to ISEF (the school did not choose me). At the state level, my research got me a high first-place award and I was elated. I also received the Naval Award for research that year. Over the summer, I was in close contact with the state science fair director and was asked to present my project at the annual STEM week where MA politicians and scientists celebrate the developments of STEM in MA. Late into 2020, I finally got an email about the opportunity to present at the MA Junior Academy of Science Symposium (MJAS) at MIT. Thrilled by another opportunity to present my research, I attended and got to see many other fantastic projects. I was then elected as a delegate to the AJAS symposium. One year later, I competed again at the state fair and attended MJAS. This will be my second year presenting at the AJAS Symposium. It has been a crazy journey, but I hope to continue my research in college. I am now developing the same project with visually impaired individuals to include more features into my tech advanced cane and create an even more accessible device for the visually impaired.