cs.HCOct 31, 2026

From White Canes to Edge AI: A Survey of Assistive Navigation Technologies for Blind and Low-Vision Individuals

Authors: Abtin Zandi, Arian Ashrafi, Sarvin Nami, Seyed Sina Negahban, Amirabbas Entezari, Parsa Abbasian, Ariyan Azami, Parisa Taherzadeh, 

Organizations: Department of Computer Engineering, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran · University of Toronto, Toronto, Canada · York University, Toronto, Canada · Department of Mathematics and Informatics, Al-Khwarizmi University, Urgench, 220100, Uzbekistan

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Independent mobility is a foundational requirement for the autonomy of blind and low-vision (BLV) individuals, yet the assistive technologies that support it have historically evolved slowly. This survey systematically reviews 108 peer-reviewed studies published between 2008 and 2026, organizing the field into four technological eras: (i) a foundational era of passive mechanical aids, namely the white cane and the guide dog; (ii) a hardware era of electronic travel aids built on ultrasonic, infrared, and satellite positioning sensors that alert users to proximity but lack semantic understanding; (iii) a deep learning era in which convolutional and transformer-based detectors introduced context-aware scene understanding, initially tethered to cloud infrastructure; and (iv) a contemporary edge-native era defined by model compression, quantization, hardware acceleration, and multi-sensor fusion that relocate intelligence onto resource-constrained wearable and mobile devices. For each era we analyze representative systems, sensing modalities, computational architectures, and reported performance, and we consolidate the field's model-optimization techniques into a single technical reference spanning four detector families and twelve compression strategies. Across the corpus, we find a persistent mismatch between rapid algorithmic progress and slow real-world adoption, driven by cognitive overload in user interfaces, fragmented evaluation protocols, and the scarcity of studies involving BLV participants. We conclude with open challenges, including seamless indoor-outdoor transition, standardized benchmarks, privacy-preserving perception, and the emerging role of multimodal foundation models, and outline a roadmap for socially acceptable, user-centered assistive navigation.

Figures & tables

Figure 1: PRISMA flow diagram of the study selection process [ 18 ] , from initial identification through to the 108 studies synthesized in this survey.
Figure 2: The four-era genealogy of BLV assistive navigation technologies used to organize this survey. Each era is defined by its dominant sensing substrate, computational paradigm, and connectivity assumption; representative systems for each era are compared in Table 4.5 .
Figure 3: Pham et al. ’s Kinect-based pipeline: an RGB-D camera feeds obstacle detection via RANSAC floor-plane fitting, with output routed to audio or a sensory-substitution device. Redrawn after Pham et al. [ 42 ] .
Figure 4: Evolution of the four object-detection families used across the corpus, positioned by publication year (x-axis, 2014–2025). YOLOv8 (via [ 57 , 23 , 32 ] and others), R-CNN (as a comparison baseline [ 1 ] ), SSD/MobileNet-SSD [ 58 , 52 , 59 ] , and DETR [ 56 ] are the specific versions with direct assistive-system adoptions documented in this survey; the remaining nodes mark the family’s technical lineage. Edge-mobility verdicts for each family are summarized in Table 4.3.5 .
\tablecaptionfont \tablenumfont Table 1. Representative assistive navigation systems across the four eras. US = ultrasonic; TTS = text-to-speech; RPi = Raspberry Pi; “BLV study” = evaluation with blind or low-vision participants; n.r. = no such evaluation reported in the source consulted. The column was assigned by a full-text keyword search of each source for participant- and user-study-related language. Five of the 22 systems listed report testing with confirmed BLV participants, and one further system reports a single-participant pilot of unconfirmed BLV status.
SystemEraYearSensingModelPlatform / feedbackReported performanceBLV study
Das et al. [ 40 ]II2020US + PIR canethresholdingcane; haptic+audio400 cm range, velocity sensingn.r.
Kasabekar et al. [ 39 ]II20243 ×\times US + GPSthresholdingmodular wearable; haptic+audio+SMS10–300 cm rangen.r.
Chintu et al. [ 41 ]II2026US + GPS (IoT)thresholdingwearable; vibration+audioreal-time proximity + trackingn.r.
Pham et al. [ 42 ]II/III2016Kinect RGB-DRANSAC geometryPC; audiohigh wall/floor-object acc.n.r.
Bai et al. [ 107 ]III2017stereo helmet camcloud SLAM + CNN + OCRcloud; audio1000+ classesn.r.
Table 5
More in this topic

Assistive Technology for Blind and Low-Vision People

See all 35 papers

Explore similar work

CardsList

  1. An Affordable AI-Integrated Smart Cane for Multimodal Mobility Assistance of Visually Impaired Users

    Sep 12, 2026Ali Akarma, Adeel Ahmad, Toqeer Ali SyedVisually Impaired Navigation AssistanceAssistive Robotics

    AI Center, Faculty of Computer and Information Systems, Islamic University of Madinah, Madinah 42351, Saudi Arabia · AI V&V Lab, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia

  2. VISTA: A Multimodal Egocentric Benchmark for Goal-Oriented Perception and Assistance to Blind and Low-Vision Users

    Oct 16, 2026Noé Mosca, Mohamed Rayan Barhdadi, Muhammad Hamza +11Assistive Technology for Blind and Low-Vision PeopleVisually Impaired Navigation Assistance

    Harvard AI and Robotics Lab, Harvard University · Télécom Paris, Institut Polytechnique de Paris · Texas A&M University +2

  3. Touvigation: Embodied Adaptive Object Acquisition for Blind and Low-Vision Users in Unfamiliar Indoor Environments

    Sep 18, 2026George Xi Wang, Xiangyu Li, Shaoyue Wen +8Assistive Technology for Blind and Low-Vision PeopleVisually Impaired Navigation Assistance

    Stony Brook University, United States · New York University, United States · Brown University, United States +6