Passenger Behavior Rating for Offline Transit Risk Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional account-based transit systems rely on transient data for risk management, which becomes outdated when fare media readers go offline, increasing fraud exposure.
Innovation Solution
Implementing passenger behavior rating systems that generate risk scores based on usage history, allowing for intelligent decision-making even when transit access readers are offline, by characterizing individuals and providing risk scores to fare media, thereby reducing the need for frequent list updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional account-based systems use transient data from back office for access decisions, then the system can make binary access decisions, but the risk management becomes outdated when fare media readers go offline
Solution Approach 1:
The system pre-calculates and stores risk scores for passengers based on their historical behavior data before they need to access the transit system. These pre-computed risk assessments are stored in the fare media itself, allowing readers to make reliable access decisions even when offline without needing to query the back office.
Solution Approach 2:
The patent introduces risk scores as an intermediary metric between passenger behavior and access decisions. Instead of relying solely on binary allow/deny lists, the system uses continuous risk scoring that can be pre-computed and stored locally, serving as a mediator that enables accurate risk assessment independent of real-time back office connectivity.
2Reliability
If the system maintains accurate positive and negative lists for fraud prevention, then fraud exposure is reduced, but frequent list updates are required when readers are online
Solution Approach 1:
The system performs preliminary risk assessment by analyzing passenger behavior history and pre-computing risk scores before the passenger attempts to access the system. These scores are stored in the fare media, eliminating the need for frequent updates of positive and negative lists and reducing the time required to maintain accurate fraud prevention data.
Solution Approach 2:
The patent transforms the static binary state of positive/negative lists into a dynamic continuous risk score parameter. This parameter can be updated less frequently since it captures nuanced behavioral patterns, and it can be stored compactly in fare media, reducing the overhead of list maintenance and updates.
3Productivity
If fare media readers operate offline, then system availability is maintained, but the exposure to fraud increases due to outdated lists
Solution Approach 1:
The patent extracts the essential risk management capability from the centralized back office system and embeds it directly into the distributed fare media readers. By storing pre-computed risk scores locally in the fare media and reader devices, the system removes the dependency on real-time back office connectivity for fraud detection, enabling offline operation without increased fraud exposure.
Solution Approach 2:
The fare media readers are empowered to make independent access decisions using locally stored risk scores without needing to query the back office. Each reader serves itself by evaluating the risk score associated with each passenger's fare media, enabling autonomous fraud prevention that maintains system availability even when offline.
Data Source
AI summary
A communication is received from a fare media which includes a token associated with a transit account and a risk score based on a usage history of the account which is indicative of a likelihood of default associated with the account. One or more positive or negative lists are received from a server. Positive lists include first account identifiers associated with accounts in good standing. Negative lists include account identifiers associated with accounts in poor standing. A determination whether the token is associated with an account represented by one of the first or one of the second account identifiers is made. A determination, based on the risk score or the determination of the association of the token is made whether to grant access to the transit system. Information related to the determination of whether to grant access is communicated to the server. Access is granted to the transit system.