Bus Stop Asset Monitoring

Transit agencies face persistent challenges in monitoring bus stops and maintaining their assets. This responsibility becomes increasingly difficult as the network of bus stops grows, leading to the impracticality of a perpetual inspection of the stops in urban environments. Traditional manual inspections are labor-intensive, infrequent, and unable to provide real-time data needed for proactive maintenance. Our innovative solution aims to shift this paradigm by turning public buses into mobile inspectors enabled by a sophisticated combination of IoT, computer vision, and generative AI technology.

The system allows in-service buses to continuously observe and catalogue asset conditions without requiring additional supervision, inspection vehicles, or personnel. As buses traverse their normal routes, the system intelligently identifies bus stops and captures visual data to help detect existing assets and evaluate their status. This enables a constant inspection cycle that promises to increase monitoring frequency and to reduce costs by allowing agencies to implement more targeted asset investment planning.

The core technology of the solution is implemented by the integration of IoT sensors, computer vision, and generative AI. Visual data is collected via the IoT device, transmitted to a cloud server where it is analyzed by state-of-the-art LLM to provide real-time description of the present assets and an assessment of the conditions such as good, damaged, etc. This generates actionable on-the-go insights for transit agencies about shelters, benches, signage, and other critical infrastructure. Our system provides access to these insights through an intuitive web application with interactive maps, dashboards, chart visualizations of statistical measures of interest, and easy management of historical data of assets and stops. All of the discussed components aim to paint an accurate up-to-date picture for transit agencies of what is out there. An intricate tracking and understanding of assets and conditions on the ground highlights maintenance priorities before impacting service quality or passenger safety.

The system opens the door for transit agencies to explore the broader implications of how they can leverage AI to dramatically improve infrastructure monitoring processes beyond the immediate maintenance benefits.