60% Less Wait – How Cincoze DI-1000 Harnesses Machine Learning to Reduce Traffic Congestion
Introduction
Everyone likes to catch the green light. Nobody likes to sit in traffic jams. However, beyond the general frustration of endless waiting, transportation inefficiencies can seriously disadvantage a city or country’s economy and environment. According to INRIX’s 2019 Global Traffic Scorecard, the American Transportation Research Institute estimates that the total cost of traffic congestion in the freight sector is as high as $74.1 billion annually. But it’s not just the economy that’s affected. Petroleum consumption, CO2 emissions, and climate change caused by traffic congestion take a toll on our planet. Now, by leveraging new technology, intelligent traffic management smooths and optimizes vehicle flow, getting people and things to where they’re going quicker than ever before.
Our customer, a research institute of environmental engineering in a European university, was tasked with this challenge when their city’s Urban Congestion Commission called for improved traffic signal control to reduce congestion and travel time. The research institute developed its own algorithm and created a new method for managing signal-controlled intersections by adopting machine learning. Their strategy uses radar to detect and then predict the behavior of traffic. With the integration of machine learning, optimized traffic and control light timings reduce red lights when there is no oncoming traffic. An embedded PC solution was required for this application to process data and connect to the peripherals.
Customer’s Requirements
Small yet Powerful
The embedded PC connects to the radar, collects traffic data, processes the data, then transmits it to the central control room. To meet t
he roadside edge computing system’s requirements, the embedded PC must have high-performance computing power in a small form factor to permit installation in the small roadside cabinet.
Multiple LAN and I/O Connection
The embedded PC must connect to the traffic management center. A network connection facilitates this communication, with the control room analyzing the incoming data, and the embedded PC commanding and controlling the traffic lights. Therefore, LAN, I/O connections, and expansion are necessary.
Outdoor Conditions
A traffic system is a harsh installation environment. The embedded PC must adapt to accommodate varying outdoor application environments, including shock, vibration, and wide temperature ranges, to ensure uninterrupted traffic service.
