Advancing Traffic Management in Bhutan: eDruk’s Integration of Machine Learning and Computer Vision

Machine Learning (ML) and Computer Vision (CV) are pivotal technologies driving modern intelligent systems. ML involves algorithms that enable computers to learn from data, improving performance on tasks like prediction and classification without explicit programming. CV, a subset of ML, focuses on enabling machines to interpret and understand visual information from the world, such as images or videos, mimicking human vision. Together, they power applications in various domains, including traffic management, where they analyze real-time data to optimize flow, enhance safety, and reduce congestion.

Applications in Traffic Systems

In traffic systems, ML and CV are used for:

• Number Plate Recognition (ANPR): CV algorithms detect and read license plates using optical character recognition (OCR), while ML models like YOLO (You Only Look Once) classify vehicles and handle variations in lighting or angles. This enables automated enforcement, such as identifying stolen vehicles or toll collection.
• Vehicle Counting: CV processes video feeds to detect and track vehicles, with ML algorithms counting them in real-time to monitor traffic density and predict jams.
• Parking Space Allocation: Systems use overhead cameras analyzed by CV to identify empty spots, and ML predicts availability based on patterns, guiding drivers via apps or signs to reduce search time and emissions.

Traffic Management in Bhutan

In Bhutan, where urbanization is increasing vehicle numbers, these technologies are crucial for sustainable transport. eDruk Private Limited, a pioneering Bhutanese IT firm established in 2004, is at the forefront of this innovation. With expertise in AI/ML, IoT, and infrastructure projects like the DrukREN network expansions and Government Intranet revamp, eDruk is building advanced traffic management systems tailored to local needs. Leveraging partnerships with the Road Safety and Transport Authority (RSTA) and Royal Bhutan Police (RBP), eDruk integrates ML and CV into solutions that enhance road safety and efficiency.

eDruk’s Traffic Management System

eDruk’s traffic management system builds on their existing projects, such as real-time GPS vehicle tracking for RSTA public buses and Bank of Bhutan fleets, as well as fiber-optic CCTV networks for RBP national security. Extending this, eDruk employs CV-powered cameras for ANPR, achieving high accuracy in detecting Bhutanese license plates even in challenging Himalayan conditions. For instance, their system uses ML models to process video streams, identifying vehicles in real-time and flagging violations like speeding or unauthorized entry.

This aligns with Bhutan’s Safe City Project, where AI-driven ANPR checkpoints monitor traffic flow and alert authorities to blacklisted vehicles.

For vehicle counting, eDruk’s system analyzes CCTV feeds with CV algorithms to tally vehicles at intersections, providing data for dynamic traffic light control and congestion alerts. Parking allocation is another key feature: Using overhead cameras and ML predictive models, the system detects vacant spots in real-time and allocates them via mobile apps like mRSTA, reducing urban parking chaos in cities like Thimphu.

This “Made in Bhutan” approach ensures reliability, with local talent from institutions like the College of Science and Technology contributing to custom ML models trained on Bhutanese datasets for better accuracy (e.g., 98.5% vehicle detection in similar local systems).

By combining these elements, eDruk’s system promotes digital inclusion and sustainability, drawing from their 20+ years of G2C projects like LabourNet and Job Portal. As Bhutan advances towards smart cities, eDruk’s innovations promise safer, smarter roads. For more details, visit eDruk’s website.