Nvidia’s accelerated data processing has helped India manage its toll traffic, which stretches over 6.4 million kilometers across 1,000 toll plazas.
The Indian road network is the second largest in the world and is largely operated manually. Traditional toll plazas, wherever they are used in the world, can cause massive traffic delays, long commute times and severe traffic congestion.
To support the automation of toll plazas across India, Indian-American technology company Calsoft helped a client implement a wide range of Nvidia technologies that were integrated into the country’s dominant payment system, known as the Unified Payments Interface (UPI).
Manual toll booths require more time and labor than automatic ones. However, automating India’s toll systems brings with it an additional difficulty: the wide variety of license plates.
Join GamesBeat Next!
GamesBeat Next brings together the next generation of leading video gamers. And you can be there on October 28th and 29th in San Francisco! Take advantage of our buy one ticket, get one free offer. The sale ends this Friday, August 16th. Become a member by registering here.
India’s non-standardized license plates pose a significant challenge to the accuracy of Automatic Number Plate Recognition (ANPR). Any implementation would need to account for these plate variations, which include different colors, sizes, fonts and placement on vehicles, as well as many different languages.
The solution developed by Calsoft automatically reads the license plates of passing vehicles and debits the respective driver’s UPI account. This approach reduces the need for manual toll collection and is a big step towards solving the traffic problem in the region.
Automation in action
This solution has been rolled out in several leading metropolitan cities as part of a pilot program. The solution offers approximately 95% accuracy in reading license plates by using an ANPR pipeline that detects and classifies the license plates as they pass through the toll plazas.
According to Vipin Shankar, senior vice president of technology at Calsoft, Nvidia’s technology was crucial in this endeavor. “Detection was particularly difficult at night.
Another challenge was improving model accuracy when pixel distortions occur due to environmental conditions such as fog, heavy rain, reflections from bright sunshine, dusty winds and more,” he said.
The solution uses Nvidia Metropolis to track and detect vehicles throughout the process. Metropolis is an application framework, set of developer tools, and partner ecosystem that brings together visual data and AI to improve operational efficiency and safety across a range of industries.
Calsoft engineers used Nvidia Triton to deploy and manage their AI models. The team also used the Nvidia DeepStream software development kit to create a real-time streaming platform. This was key to efficiently processing and analyzing data streams and incorporating advanced features such as real-time object detection and classification.
Calsoft uses Nvidia hardware in its AI solutions, including Nvidia Jetson Edge AI modules and Nvidia A100 Tensor Core GPUs. Calsoft’s toll booth solution is also scalable, meaning it is designed to accommodate future growth and expansion needs and can better ensure sustainable performance and adaptability to changing traffic conditions.
