Leezenflow: Green wave for cycle traffic

Iterative design and implementation process

The collaboration between city administration, technical implementation partners, and the local college and university was essential to the success of the iterative design and implementation process. A comprehensive test phase, including rigorous evaluation, ensured that conditions for the target group – cyclists – were genuinely improved without introducing additional safety risks. The rollout of Leezenflow at nine additional locations along the city’s cycling routes built on insights gained from the prototype testing phase. These insights led to key optimizations, primarily in display improvements and the integration of vehicle-to-everything (V2X) technology, which enables communication between traffic lights and vehicles. This advancement supports more accurate predictions and more compact hardware solutions.

Utilisation of existing data

Leveraging existing data on average cycling speeds across various routes, gathered from previous city cycling projects, allows for more accurate predictions of cyclists’ ability to reach green lights. A critical aspect of this process is the synchronization of traffic light data with Leezenflow information. By integrating data from both sources, traffic light timing can be managed more efficiently, optimizing traffic flow. 
 

Non-proprietary software and hardware components

The choice of non-proprietary software and hardware components allows for easy transferability of the solution. The international OCITv3 standard is applied, enabling the base software to be used immediately where compatible hardware is available. The Leezenflow devices’ enclosure and hardware can also be assembled independently according to the provided detailed instructions.

Online component description and documentation for replication 

Comprehensive documentation on the enclosure design, hardware components, assembly time, and source code ensures precise traceability of all solution components and facilitates easy transferability. Documentation is accessible online (available at: https://github.com/bCyberGmbH/leezenflow-doku/). The solution has already been adopted by CityLAB Berlin in the "Kottiflow" project.
 

Measuring effectiveness and optimisation through analyses

Ongoing evaluation of the solution is key to ensuring its continuity. Regular feedback and analysis not only confirm effectiveness but also provide opportunities for targeted optimizations, such as technical improvements to meet user needs.

Comparison of actual and target data

Collected actual and target data from the traffic light control system will enable more precise management of bicycle traffic in the future. Through vehicle-to-everything (V2X) technology, further optimizations are possible in the medium term - through targeted reactions of the traffic light control to bicycle traffic.

Continuation within existing traffic light operations

The continuous operation of the Leezenflow devices is secured as part of the existing traffic light system. However, potential challenges, such as vandalism or weather-related damage, should be considered in planning, as they may lead to increased costs due to equipment repairs or replacement.
 

Local challenges

Münster, a significant academic center due to its multiple universities, is geographically positioned between the Ruhr area and Osnabrück. Known as a bicycle-friendly city, it has a cycling mode share of over 40% of total mobility. However, its location also brings a high volume of commuter traffic from surrounding towns, much of which is motorized transport. 

In light of climate-friendly and sustainable mobility goals, it is essential to reduce the the high proportion of motorized transport and expand alternative modes of transportion. Smart traffic management solutions help make these alternatives more efficient, safer, and more appealing, with cyclists as a key target group. By enhancing speed, comfort, and safety, innovations like Leezenflow can significantly improve central cycling routes and increase the attractiveness of cycling.

Planning goals

Münster’s Smart City strategy aims to enhance quality of life and road safety through digital technologies. In the ‘mobility and transport’ sector, achieving a balanced mix of transportation modes is essential for improved air quality and reduced noise.

Leezenflow’s implementation on Münster's main cycling routes contributes to optimized safety and improved connectivity, extending its benefits to cyclists and commuters alike. This, in turn, increases cycling’s appeal and supports the city’s commuter traffic goals.

Approach to measuring impact

The project’s success is measured by the installation and functionality of the Leezenflow devices. Additionally, an impact assessment evaluates progress toward project targets. This assessment includes a survey to gauge public acceptance of Leezenflow, focusing on user engagement, behavior changes (e.g., speed adjustments due to Leezenflow), with results published in Q3 2024.

Further parameters for a comprehensive impact assessment, aligned with planning objectives, may include:

• Improvements in perceived cycling friendliness
• Changes in cycling frequency along Leezenflow-equipped routes
• Air quality and noise level improvements along Leezenflow routes.

Process steps

Leezenflow was developed through an agile, iterative process. The initial concept originated at the "Münsterhack 2019" hackathon. Münster’s city administration took up the idea, implementing a prototype with support from the Federal Ministry of Transport and Digital Infrastructure. The first Leezenflow display was tested on Münster’s Promenade in May 2021, with evaluations based on citizen surveys and on-site traffic observations and measurements.

As part of the Model Project Smart Cities funding, Münster expanded and refined Leezenflow based on test results through the following steps:

1. Identification of nine additional locations for Leezenflow devices (three on the Promenade and six on main cycling routes within the city).

2. Conceptualisation for further development of Leezenflow, including: 

• New LED visualisation, incorporating a yellow phase for better prediction.
• Inclusion of average road speeds to improve predictions for green phase reach. 
• Integration of a roadside unit for traffic light communication within the display.
• Installation of vehicle-to-everything (V2X) antennas.
• Addition of sensors to measure temperature, humidity, air pressure, and input voltage.
• Installation of LTE routers for remote device access and monitoring.

3. Production and installation of nine additional Leezenflow devices.

4. Upgrade of the existing Leezenflow device on the Promenade.

5. Functional testing of all devices.

6. Ongoing evaluation of the expanded system.

7. Creation of an explanatory video and demonstration device for public communication.

Governance

Within Münster’s organisational structure, the Smart City staff unit coordinates the initiative. This unit collaborates with the city’s Office for "Mobility and Civil Engineering" ("Department for Traffic Management and Smart Mobility"), the "City Planning Department" ("Department for Historic Building Protection"), and the Office for "Press and Public Relations".

Implementation involves multiple partners: the University of Applied Sciences Münster (enclosure development), the University of Münster (prototype evaluation), engineering office nts (site investigations and display enhancements) as well as the companies Swarco Ampeltechnik, and bCyber GmbH (Leezenflow programming and production), Gipfelgold GmbH (production of explanatory video) and the Code for Münster initiative. The public order office and local police support equipment installation.

Costs for acquisition and operation

The total costs of €300,000 are to be understood as an upper limit and were required for the installation of 10 devices in Münster (installation, test operation, evaluation and start of stabilization). 

Participation and communication

The initial Leezenflow concept emerged from the grassroots "Münsterhack 2019" hackathon, reflecting the needs of Münster’s cycling community. Evaluation during prototype testing involved citizens to confirm both functionality and safety, with results indicating that most users could optimize their cycling behavior.

Device locations were selected through a local political process in consultation with district representatives. Effective communication, through social media, local media (Westphalian News), and an explanatory video, is essential to enhance public understanding and acceptance of the technology.

Technical infrastructure

Leezenflow is an open-source solution that can be implemented independently with the necessary hardware, without proprietary restrictions. All project data is accessible in a GitHub repository under the Creative Commons BY-NC-SA 4.0 license:GitHub repository.

The code, written in Python, manages key functions: reading traffic light data from control units, predicting green-phase reach based on average cycling speed, and displaying the prediction via Leezenflow. Code can be accessed here:Leezenflow code repository.

Detailed assembly instructions for recommended components, along with required hardware for the Leezenflow display, are available: Assembly documentation.

Each Leezenflow display includes an enclosure and core hardware components: a mini processor (Raspberry Pi, Linux OS), memory card, temperature sensor, input voltage sensor, sensors for air pressure and humidity, LTE router, GNSS (satellite navigation) antenna, Cellular-Vehicle-to-Everything (C-V2X) antenna, and LED panels.

Each device requires an initial setup, with programming adjusted for its distance from the traffic light. 

Data basis

Data collected from Münster’s previous cycling projects, such as "Stadtradeln" and "Dein AppGrade", has refined the accuracy of predictions for cyclists’ ability to reach green lights by factoring in average cycling speeds on busy routes. With the Leezenflow system in place, real-time data from the traffic lights (via SPaT telegrams) is now transmitted several times per second using V2X technology under the OCIT standard. This data will allow more precise forecasting for cyclist-specific applications in future.