Ultrasonic technologies versus Artificial Intelligence in canes for people with visual impairments: A systematic review (#871)

Authors

Flores Masías, Edward José

Soto-Palomino, Wilfredo

Palomino-Peralta, Carmen

Cuba-Aguilar, Cesar-Raul

Abstract

Abstract– This study examines the smart canes that have emerged in recent years to support people with visual impairments. The study aimed to evaluate and compare the effectiveness of three main technologies in smart canes for people with visual impairments: ultrasonic, artificial intelligence (AI), and hybrid systems. The methodology involved a systematic review of literature published between 2020 and 2025. Eight independent searches were conducted in four databases, identifying 908 initial articles, which were reduced to 180 unique articles after duplicate removal. The following dimensions were evaluated: accuracy, detection range, user satisfaction, autonomy, response speed, and environmental robustness. The results showed that hybrid systems obtained the highest overall score (89.2/100), with 95% accuracy, a range of 0.3–9 m, and 93.5% user satisfaction. Ultrasonic systems achieved a score of 87.4/100, standing out for their low cost. Systems based solely on AI scored 80.1/100, limited by shorter autonomy and vulnerability in low-light conditions (60-65% performance). The study concluded that hybrid systems represent the state of the art in overall effectiveness, but ultrasonic systems demonstrate surprising competitiveness. There is no single optimal solution; the choice depends on the context of use, budget, and specific needs. Critical gaps were identified in longitudinal studies, evaluations in diverse real-world contexts, and rigorous economic analyses.