This paper presents an assistive navigation approach for people with visual impairments (PwVI), based on the modeling of an accessibility heuristic and the prototyping of an application that integrates environmental information into routes generated by a pedestrian-oriented routing API. The objective was to improve the characterization of pedestrian mobility conditions, addressing limitations observed in existing navigation solutions. The methodology included in loco mapping of urban barriers and sensory reference items, evaluation of navigation applications and routing APIs, development of the heuristic, and its partial implementation in a functional prototype. Results indicated a predominance of barriers in the urban infrastructure, limited sensitivity of the evaluated systems to accessibility conditions, and insufficient navigation instructions for the needs of PwVI. Preliminary application of the heuristic demonstrated its feasibility to enrich routes with relevant attributes and differentiate trajectories according to their accessibility conditions. The findings suggest that the proposed approach contributes to more inclusive navigation systems and highlights the need for future work involving full heuristic implementation, field testing, and user-centered evaluations.