Transurban Adopts AI Technology for Improved Toll Collection

Transurban, a major operator of toll roads across Australia, has successfully integrated artificial intelligence (AI) to streamline its billing process for users whose e-tags are not detected or recognized. The company has developed an advanced “auto-correction” AI model designed to address instances where its electronic toll collection system fails to identify vehicles.

This model leverages advanced artificial intelligence to process images captured by automatic license plate recognition (ALPR) cameras. It meticulously analyzes each image to extract essential details, including the vehicle's make, number plate, and precise location. Once this information is gathered, it is cross-referenced with a comprehensive database of customer information to ensure accurate invoicing. By integrating this data, the system effectively identifies vehicles that were missed by the e-tag system, ensuring that every road user is correctly billed. This automated approach enhances the accuracy of billing and reduces the need for manual review, streamlining the invoicing process and improving operational efficiency.

The introduction of this AI model has led to a substantial improvement in efficiency, reducing the number of images requiring manual human review by up to 40 percent, according to Artak Amirbekyan, Transurban’s head of data, AI, and machine learning. Amirbekyan highlighted this development during his presentation at the Gartner Symposium held in Gold Coast. He explained that while the majority of customers use e-tags successfully, there are occasional failures due to malfunctioning tags, user errors, or other issues. In such cases, accurate license plate recognition becomes crucial to ensure proper billing.

Transurban processes around 2.5 million customer trips daily across its network of toll roads. For trips where the e-tag is not detected, ALPR cameras capture images of the vehicles. When these images are challenging to read or unclear, they are forwarded to Transurban’s team of agents who manually review and process them to determine the vehicle’s identity and billing details.

The AI-powered ALPR system, which operates on Amazon SageMaker, achieves an impressive 99 percent accuracy rate in identifying vehicles. This high level of precision is essential for avoiding errors and minimizing customer complaints related to incorrect billing. The system’s effectiveness is a testament to Transurban’s commitment to leveraging advanced technologies to improve operational efficiency.

In addition to the AI-enhanced ALPR system, Transurban’s data, AI, and machine learning team is exploring other innovative applications of AI technology. These include enhancing road safety, detecting incidents, and optimizing tunnel ventilation systems. The company’s dedication to AI is clearly reflected in its workforce, with nearly 40 percent of its employees engaged in technology-related roles. This significant focus on tech talent underscores the company's commitment to leveraging advanced technologies and data-driven solutions to drive innovation and maintain leadership in the transportation sector.

CTO Tanya Trott, also speaking at the symposium, underscored that despite Transurban’s status as one of the safest road operators globally, there remains considerable potential to further improve safety through the strategic use of data and AI. The company’s continuous integration of AI across its operations highlights a broader commitment to pushing the boundaries of innovation and excellence in transportation management. By harnessing advanced technologies, Transurban aims to enhance operational efficiency, improve safety, and deliver a superior experience for its customers. This proactive approach not only addresses current challenges but also positions Transurban at the forefront of technological advancements in the transportation sector.