teLLMe Why (Ain't Nothing but a Jam): Exploratory Causal Analysis of Urban Driving Data

2026-07-16

AI Summary

This work addresses challenging causal questions in urban driving observational data—such as the effect of weather on traffic density—by proposing teLLMe, a novel system that integrates causal structure learning, the DoWhy framework, and a pattern-aware large language model. teLLMe constructs a causal graph from structured event tables using the PC algorithm, estimates causal effects via linear regression with Bootstrap-based stability validation, and automatically translates natural language queries into formal causal inquiries. The system generates interpretable “causal cards” that report estimated effects, adjustment sets, and key modeling assumptions. Experiments on the BDD traffic dataset demonstrate that teLLMe effectively uncovers plausible causal relationships among weather conditions, rush hours, and traffic density, while transparently communicating inference uncertainty and underlying assumptions.

Abstract

Traffic agencies now have access to large volumes of video-derived data for studying safety and congestion. Most of these data are observational and collected without interventions, which makes causal questions such as "How would rain change traffic density?" difficult to answer. We present teLLMe, a system for exploratory causal analysis of urban driving datasets. The system starts from a structured event table built from dashcam annotations and combines causal structure learning with the PC algorithm, bootstrap-based stability checks, and query-specific effect estimation using linear regression and DoWhy. Natural-language questions are mapped to structured causal queries through a schema-aware LLM, enabling users to specify treatments, outcomes, and subpopulations. teLLMe returns a "Causal Card" that summarizes effect estimates, adjustment sets, DAG support, and assumptions, followed by a short natural-language explanation. Case studies on BDD-derived traffic events show that the system can surface plausible relationships involving weather, peak hours, and traffic density, while making uncertainty and modeling choices explicit. The system is designed as a tool for hypothesis generation and expert reasoning rather than a source of definitive causal claims.

Problem

• causal inference
• observational data
• urban driving
• traffic analysis
• causal questions

Innovation

• causal discovery
• exploratory causal analysis
• schema-aware LLM
• Causal Card
• urban driving data