[Paper Review] AIoT-based Continuous, Contextualized, and Explainable Driving Assessment for Older Adults

Yimeng Liu, Fangwei Zhang+10 more|arXiv (Cornell University)|Feb 28, 2026

Older Adults Driving Studies0 citations

TL;DR

This paper proposes AURA, an AIoT framework for continuous, context-aware, and explainable driving assessment of older adults, using multimodal in-vehicle sensing and edge computing to distinguish aging-related changes from situational factors.

ABSTRACT

The world is undergoing a major demographic shift as older adults become a rapidly growing share of the population, creating new challenges for driving safety. In car-dependent regions such as the United States, driving remains essential for independence, access to services, and social participation. At the same time, aging can introduce gradual changes in vision, attention, reaction time, and driving control that quietly reduce safety. Today's assessment methods rely largely on infrequent clinic visits or simple screening tools, offering only a brief snapshot and failing to reflect how an older adult actually drives on the road. Our work starts from the observation that everyday driving provides a continuous record of functional ability and captures how a driver responds to traffic, navigates complex roads, and manages routine behavior. Leveraging this insight, we propose AURA, an Artificial Intelligence of Things (AIoT) framework for continuous, real-world assessment of driving safety among older adults. AURA integrates richer in-vehicle sensing, multi-scale behavioral modeling, and context-aware analysis to extract detailed indicators of driving performance from routine trips. It organizes fine-grained actions into longer behavioral trajectories and separates age-related performance changes from situational factors such as traffic, road design, or weather. By integrating sensing, modeling, and interpretation within a privacy-preserving edge architecture, AURA provides a foundation for proactive, individualized support that helps older adults drive safely. This paper outlines the design principles, challenges, and research opportunities needed to build reliable, real-world monitoring systems that promote safer aging behind the wheel.

Motivation & Objective

• Motivate safer aging in driving due to cognitive and sensory decline and rising older-driver prevalence.
• Develop a continuous, real-world driving assessment framework that captures fine-grained behaviors in naturalistic settings.
• Enable context-aware interpretation to separate age-related changes from environmental factors.
• Provide explainable AI outputs that are actionable for drivers, families, and clinicians.

Proposed method

• Introduce AURA, an AIoT framework for continuous, privacy-preserving in-vehicle sensing and edge processing.
• Use multimodal sensing to capture vehicle control, in-cabin behavior, and environmental context.
• Employ context-aware data fusion and longitudinal modeling to produce behavior trajectories and safety indicators.
• Incorporate explainable AI via causal, mechanistic interpretations and in-vehicle deployment for privacy.
• Leverage datasets (CARLA simulator, LongROAD, DRIVES) to study senior driving patterns and develop an interpretable unsupervised analysis pipeline.

Figure 1 . Overview of AURA

Experimental results

Research questions

• RQ1 How can continuous, real-world driving data be used to assess cognitive and functional changes in older drivers?
• RQ2 What contextual factors (traffic, weather, lead vehicles) modulate age-related driving behaviors?
• RQ3 How can driving signals be translated into explainable, clinically useful assessments and interventions?
• RQ4 What modeling approaches yield stable, interpretable insights rather than opaque predictions?

Key findings

• Seniors show slower brake/accelerator changes, lower speeds, and more head-scanning than younger drivers in both simulated and real-world data.
• Senior driving is highly context-dependent, with lead-vehicle presence and weather conditions significantly altering behavior patterns.
• LongROAD and DRIVES data reveal aging-related mobility reductions and strong self-regulation, including night-driving avoidance and route preferences.
• PCA and clustering indicate strong within-driver stability and diverse individual profiles, but limited separation by cognitive status across individuals.
• Explainable AI challenges are acknowledged; current Unsurvised approaches identify signatures but do not reveal underlying causes, highlighting need for causal modeling.

Figure 2 . Overview of the CARLA simulator environment employed for controlled driving data collection.