Case studySmart City

Automate Toll Payment & Reduce Losses with AI Vision

Known to be the largest in Malaysia, PLUS Malaysia Berhad uses Tapway VehicleTrack to automate payment for multi-class lanes and detect toll fraud with AI vision.

Client: PLUS Malaysia Berhad · >99.4% ANPR character accuracy

Automate Toll Payment & Reduce Losses with AI Vision

Overview

The PLUS Expressways Berhad is the largest highway concessionaire — a build–operate–transfer operator company in Malaysia. Known to be the largest in Malaysia, PLUS Malaysia Berhad is using Tapway VehicleTrack to help automate payment for multi-class lanes, where manual labor was previously required to classify each vehicle to assign the appropriate toll rate. In addition, there were cases of fraud throughout the expressway which impacted toll revenue.

Solution

Tapway VehicleTrack is deployed on the highway to count and classify vehicles, capturing:

  • Vehicle class — for automated toll-rate assignment
  • Licence plate — for fraud detection and vehicle tracking
  • Vehicle color and brand logo — additional identification layers

Tapway uses AWS IoT Greengrass to seamlessly manage all the edge servers while ensuring they have the capabilities of the cloud without the need for manual intervention. AWS IoT Core securely connects the edge servers at toll plazas to the cloud, and allows Tapway to develop and deploy applications even when they are offline — ready to be pushed once the connection is running.

Results

With >99.4% ANPR character accuracy and >97.6% vehicle classification accuracy for toll payment, Tapway VehicleTrack has enabled PLUS Malaysia to:

  • Automate toll payment rate according to vehicle class group
  • Detect fraud cases by matching card ID with license plate to determine the same card usage for both entry and exit lanes

With Tapway VehicleTrack in place, PLUS can now automate payment, reduce labor, reduce losses due to fraud, and track movement of vehicles on their premises.

Put this to work on your site

Bring us your hardest visual workflow — we'll show you what AI can see.