Abstract

Vertically complex urban environments impose elevated spatial cognitive load on pedestrians, a  demand that static wayfinding infrastructure is structurally incapable of addressing. Smart cities  currently lack a formal cognitive navigation operating layer for managing pedestrian movement in  multi-level urban systems. This research introduces and evaluates a Cognitive Digital Twin  Operating System (Cognitive OS) — a city-scale adaptive navigation infrastructure integrating  Digital Twin environmental modelling, AI-driven route optimisation, real-time crowd intelligence,  and spatially embedded adaptive guidance to predict, manage, and reduce spatial cognitive load in  vertically complex environments. The study deploys AI-mediated human behavioral persona simulation as an independent methodological  contribution. Each of the 28 study agents constitutes a decoded human naviga- tional persona: a  behavioral reconstruction calibrated to empirical wayfinding parameters from the navigation science  literature (Argyle et al., 2023; Park et al., 2023; Aher et al., 2023), follow- ing the  simulation-first validation paradigm in which AI agents serve as surrogate participants for  rigorous proof-of-concept evaluation. A within-subjects, counterbalanced design compared  performance across an Adaptive Cognitive OS (ACO) condition and a Static Signage (SS) baseline in a  fifteen-floor mixed-use virtual building. Measures included NASA-TLX cognitive workload, wrong-turn  and hesitation frequency, navigation time, path efficiency, eye-tracking gaze metrics, and  post-navigation spatial recall. Results proved large, Significant improvements across all primary outcomes. NASA-TLX composite  workload fell by 37% (d = 1.76, p < .001); wrong-turn frequency by 67% (r = 0.93, p < .001);  navigation time by 35.9% (d = 2.29, p < .001); path efficiency improved from 0.61 to 0.87 (d =  2.37, p < .001); and visual search time by 61.0% (d = 2.80, p < .001). Spatial recall showed no  statistically significant difference across conditions (d = 0.38, p = .054), a finding that remains  inconclusive given the study’s power at that effect size. These findings establish that spatial cognitive load is a measurable, manageable urban in-  frastructure variable and that the Cognitive OS provides a viable platform for its management at  city scale. Five evidence-based design principles are derived. The research makes contributions to Cognitive Load Theory, digital twin systems, and smart city infrastructure design.

Publication Date

Spring 2026

Document Type

Thesis

Student Type

Graduate

Degree Name

Professional Studies (MS)

Department, Program, or Center

Graduate Programs & Research

Advisor

Ayman Ibrahim

Comments

This thesis has been embargoed. The full-text will be available on or around 5/12/2027.

Campus

RIT Dubai

Download

Share

COinS