Beyond RFID: Exploring TNG’s Titan Flow technology at MyAsean Roads & Traffic Tech Expo

KUALA LUMPUR: Touch ‘n Go revealed an in-depth look into Titan Flow, the company’s multi-lane fast flow (MLFF) platform that uses RFID as the primary detection method, at the MyAsean Roads & Traffic Tech Expo in KL today.

The company added the technology is supported by automatic number plate recognition (ANPR), 3D LiDAR, and AI to detect vehicle information such as RFID tag numbers, model, colour, and speed up to 320kmph.

“LiDAR captures the size and 3D profile of each vehicle, and when combined with RFID and camera data, it creates a system that helps highways to prevent revenue leakages,” said Touch ‘n Go CEO Prabakaran Sangarajoo, who presented the Titan Flow tech during an interview with LifestyleTech in KL today (Nov 5).

He added that additional technologies are incorporated to support the primary detection method and maintain high accuracy under Malaysia’s varying weather conditions.

He claimed that ANPR can achieve about 94% to 99% accuracy rate while RFID has a 99.9% detection rate.

Prabakaran said the system will also be able to capture number plates and connect with the Road Transport Department to trace vehicle movements and support enforcement actions such as roadblocks. 

“Using AI analytics, the system can even flag vehicles with outstanding toll payments for further action,” he said. 

However, Prabakaran said the company needs to onboard more users for Radio-frequency identification or RFID tags for toll payments in order to start implementing the MLFF platform.

“Current RFID penetration rate stands at around 20% to 21%. We still have about 80% of users that need to be onboarded for the transition to MLFF,” said Prabakaran.

Prabakaran acknowledged that users have encountered issues with the current RFID system, such as being stuck at toll gantries when their vehicles are not detected.

“The transformation will also require more RFID lanes and an improved user experience,” he said.

Based on statistics, chief technology officer Desmond Chin Ye Foong said more than 90% of RFID users get stuck at the gantry due to insufficient balance. To counter the issue, the company introduced SOS Balance, a feature that allows users with insufficient balance to pass through the toll and perform the payment later via the Touch ‘n Go eWallet in 24-hours. 

Since it was introduced in January, Chin said the feature has helped over 14 million users who had insufficient balance, covering more than RM42mil worth of toll payments in just nine months.

“In terms of repayment, we’re proud to share that there is a 98% recovery rate within the 24-hour window. We see this as a precursor towards a possible gateless toll payment system. What if we open up the barrier?” said Chin.

Chin explained that Titan Flow, conceptualised in 2021, is currently undergoing a proof of concept in a Shah Alam township involving 55 paying test users. 

“At the test gantry, the system captures vehicle data as each car passes through. All the information is displayed on a live dashboard, providing real-time updates such as the number of RFID transactions for the day and the toll plazas with the highest activity,” Chin said, adding that it can also flag irregularities such as mismatched or unregistered vehicle data.

As the system is developed locally, Prabakaran said the company is ensuring it meets current and future regulatory requirements. 

“It includes an enforcement module, and we can provide the application to the selected enforcement agency in the future,” he added.

Prabakaran added that the system is also built on an open payment system, meaning that users are not required to pay for tolls with only Touch ‘n Go payment systems.

“It can be MyDebit, Visa or Master and even any e-wallet. It will be open to all when it’s done,” Prabakaran said.

He maintained that RFID registration is “crucial with at least 85% to 90%”.

The company is focused on improving the adoption rate with efforts including working with car manufacturers to include pre-fitted RFID tags as a feature.