Mr. Thomas Yoon
+82-10-5380-0313
This case study outlines how e-con Systems addressed the challenges of a U.S.-based tolling infrastructure provider and delivered a customized solution that delivered tangible business benefits.
About the Client
The client is a leading technology infrastructure provider in the tolling industry, focused on modernizing highway toll collection through AI-driven, low-friction solutions.
Client Requirements
The client approached us with a set of clearly defined requirements for their intelligent traffic system (ITS) deployment. Their goal was to implement a high-performance, power-efficient imaging solution capable of detecting and monitoring vehicles in real time.
Global shutter
Camera to clearly capture fast-moving vehicles
Motorized Zoom lens
To identify vehicles between 10 to 40 meters
Integrated IR cut filter
Ambient light sensor, and support for strobe output to ensure consistent day/night imaging
Dual-stream encoding
Support for dual-stream encoding (MJPEG + H.264) with advanced video encoder design for bandwidth efficiency
Low-power envelope
For camera + encoder to support solar use
Our Solution
e-con Systems delivered comprehensive hardware and AI-ready platform designed for high-accuracy, low-power, and low-friction deployments. This solution combined custom camera hardware, high-efficiency video encoding, AI software compatibility, and rugged integration with a thermal hood for roadside use in solar-powered environments.
Key Highlights of the Solution
Camera Hardware
e-con Systems recommended a Sony Pregius S high resolution Global Shutter Camera, ensuring sharp capture of fast-moving vehicles without motion blur. This was paired with integrated motorized zoom optics to cover a range of 10–40 meters, enabling accurate vehicle identification across multiple lanes.
The design included a mechanical IR cut filter, an ambient light sensor, and a strobe output for day/night adaptability. The camera was housed in a rugged IP67-rated enclosure with a thermal shielding hood, making it suitable for outdoor roadside environments.
Video Encoder Design
e-con Systems integrated an Ambarella CV72-based encoder board. This board supported dual-stream encoding (MJPEG + H.264/H.265), enabling real-time transmission while keeping network usage optimized.
Time Stamping and Synchronization
The client required timestamps with ~10ms resolution to be added as metadata on each streamed image frame, along with a reliable time synchronization method to ensure consistency across devices.
To address this, e-con Systems® implemented precise frame-level timestamping, enabling every captured image to be tied accurately to tolling events. For synchronization, the solution could work seamlessly with an NTP client, making it simple to integrate into existing infrastructure while maintaining alignment across multiple cameras.
AI-Compatible Software Platform
To enable real-time analytics, e-con Systems delivered an RTSP-based streaming software stack with full control APIs for image tuning and optics management. The platform supports edge AI inference on Raspberry Pi Compute Modules, enabling customers to run their own AI models directly at the edge without reliance on high-bandwidth cloud connectivity.
Seamless Integration
The entire vision solution is designed to work within the client’s existing Raspberry Pi + solar setup, minimizing rework while unlocking next-level capabilities:
- Real-time data capture and analytics
- Compatibility with any ETC backend
- Support for reservation-based traffic flow models
- Unlocking free-flow tolling conditions 24/7
Business Benefits
Transponder-free, Infrastructure-lite Tolling
By eliminating the need for overhead gantries or RFID transponders, the system reduced hardware costs & deployment complexity enabling easy roadside installation with minimal civil work.
Minimal Power Draw
The low-power design made the platform ideal for solar powered setups, allowing weeks not months deployment timelines, even in remote or constrained locations.
High-accuracy Vehicle Detection
The high-resolution global shutter camera and Ambarella CV72-based encoder enabled accurate capture of fast-moving vehicles, supporting >99.95% detection and classification accuracy, even in variable lighting or weather.
AI at the Edge: Smarter, Predictive Tolling
With support for customer-deployed AI models, the platform enabled:
- Accurate vehicle occupancy detection (1-4+ passengers)
- Speed and violation analysis
- Dynamic pricing based on predictive modeling
- Reservation-based lane allocation
Flexible Integration into Existing Systems
Seamless compatibility with the client’s Raspberry Pi setup and software stack helped reduce time-to-market and avoided the need to overhaul backend infrastructure.
About e-con Systems
e-con Systems® designs, develops, and manufactures OEM cameras. With 20+ years of experience in embedded vision, we focus on delivering camera solutions to industries such as retail, medical, industrial, agriculture, smart city, and more.
e-con Systems' wide portfolio of products includes Time-of-Flight cameras, MIPI camera modules, GMSL cameras, USB 3.1 Gen 1 cameras, stereo cameras, GigE cameras, low light cameras, and more. Our cameras are currently embedded in over 350+ customer products. So far, we have shipped over 2 million cameras to the United States, Europe, Japan, South Korea, and many other countries.
Employee Login
