Thales wins award for fraud detection in fare collection

The solution detects transaction-based fraud in public transport, like media tempering or misuse.

Thales has grown to become a global player with an impressive track record in the field of ground transportation systems, both in the freight and passenger transport markets, with expertise ranging from greenfield projects to upgrades and modernisation of existing infrastructure.

It has recently developed a solution to detect transaction-based fraud in public transport, like media tempering or misuse. The solution combines a robust rule-based engine for detecting known anomalies and an AI (artificial intelligence) model to identify previously unseen irregularities. It is designed to be future-proof and adaptable, and it has been tested on real-world data, exceeding the objectives and targets set forth at the beginning of the project.

The rule-based engine provides a simple and interpretable approach for anomaly detection, implementing known abnormal pattern detection, for example, a passenger requesting too many refunds or travelling at light speed between two stations.

Meanwhile, the unsupervised AI model uses the data to detect any divergence from typical behaviours learned from millions of patterns. The specificity of the model is that it does not require human-crafted features, which can be time-consuming and prone to biases, which is a common issue in many anomaly detection methods.

This makes it a unique and innovative approach to anomaly detection, allowing the model to learn from the data in its most natural form and detect anomalies that may not have been apparent at the time of model creation. The combination of these two approaches provides a comprehensive and effective solution for detecting both known and unknown anomalies.

The solution capacity allowed users to analyse in real-time more than 20 million transactions per day. It also resulted in the development of a dashboard to monitor the detected anomalies and enable analysis by an operator for decision and action.

For this development, the company was given the AI – Transportation award in the recently concluded Hong Kong Business Technology Excellence Awards. Now in its fourth year, the awards programme recognises outstanding companies that have made exceptional contributions in pursuit of technological innovation.

In selecting the winners, it crowned the best first-rate firms with the most innovative and unique IT initiatives and solutions that helped push forward doing business in Hong Kong.

The company was awarded for its efforts to allow customers to stay ahead of fraudsters. The approach generated significant interest from operators facing fraud challenges. Overall, the proposed idea of anomaly detection in fare collection systems is a unique, effective, and dynamic solution to detecting certain types of fraud in public transport.

The Hong Kong Business Technology Excellence Awards is presented by Hong Kong Business Magazine. To view the full list of winners, click here. If you want to join the 2024 awards programme and be acclaimed for your company's outstanding contributions in pursuit of technological innovation, please contact Julie Anne Nuñez at [email protected].