Council’s objectives were to improve traffic flow and reduce brake/acceleration patterns, contributing to reduction of its carbon emissions (target:20% by 2020) to meet EU 'Horizon 2020' goals. In December 2015, the Council Department of Transportation and [ui!] began a joint project to create the city’s first traffic app and open data platform for traffic data. The project has three components:
- Traffic signal phase prediction
- Information sharing
- An open data portal
For all of these components the collection and dissemination of traffic data is at the core of our approach, as shown in the diagram below:
TRAFFIC LIGHT ASSIST
[ui!]’s UrbanPulse open data platform has been used as the basis for a Traffic Light Assist system. In this project, [ui!] UrbanPulse integrates with the city’s traffic control system to predict short-term traffic light phase changes, and shares this information with vehicles. The solution allows the adaptive cruise control in the vehicle to adapt the car's driving pattern to minimise fuel usage and stopping time at lights.
[ui!] delivers traffic data to citizens, business and the automotive industry and is the foundation for new digital services for transport disruption and creates unique conditions for eco-driving through driver assistance systems.
The free traffic-app for citizens provides:
- Qualitative visualization of traffic conditions
- High degree of updating, which is particularly crucial in the inner city area
- Construction Site information (both current and planned)
- Optimized for smartphones
The information is also shared with Continental AG who then provides it directly to vehicles.
Some of the data collected by [ui!] UrbanPulse is also made available as open data. This data then be used as a basis for citizen services, and support the private sector and researchers in development of innovative transport applications
The real-time open data portal provides:
- Aggregated detector readings
- Historical data
- Signal location plans and translation tables
- Simple data format
- Open for everyone
The decision of what data to release through the open data portal, and when, is made by the Council.
The open data portal is designed to be scalable so the city can expand its services by including any other urban data from real-time sensor data to real- time and historical data from any enterprise management system and data from 3rd parties such as weather forecasting sites.
Technical outcomes of the project include:
- A new parameterized forecasting method for traffic signal phase changes with very low latency
- Creation of a proven algorithm for traffic light prediction
- Demonstration on Test Track with Web Application
- A very low latency (0.6s. 1.1s) achieved by leveraging Microsoft Azure cloud technology and fast
- Scale-up to cover the entire city area
- Communication to end devices over SSL / TLS
- Client-side processing
- Data anonymized and not re-identifiable