The DEV-D (Dynamic Exposure Visualisation Dashboard) underwent extensive refinement and debugging during field testing phases in Flanders and Berlin.
Over 750 kilometres were covered - primarily on foot and by bicycle - during the closed testing round to optimise the initial versions of both DEV-D and DEVA (Dynamic Exposure Visualisation App).
These efforts were further extended during the open testing round, during which additional debugging and fine-tuning were performed, informed by feedback from over 970 recorded trips. This open round involved a diverse array of stakeholders, including:
Bicycle communities;
School children and educators after presentations and demonstrations in educational institutions;
Attendees at high-profile events, such as the VeloCity Conference (Ghent), where the DEVA-DEV-D integration was showcased;
Volunteering citizens, including local champions;
Subject-matter experts;
Civil servants.
Significant technical improvements were implemented at multiple levels
1. DEVA trip recorder and data manager:
Trip geometry refinement: The geometry of recorded trips was fine-tuned to manage outliers, thereby improving data granularity and accuracy.
Recording frequency adjustments: The recording frequency was adapted to suit different modes of travel (such as faster-moving vehicles), taking into account sensor limitations.
Sensor identification: The functionality for sensor ID scanning was enhanced.
Battery and bandwidth optimisation: Efficiency was improved through better management of power consumption and data bandwidth. Features like server connection optimisation, bundled data upload, and post-trip data transmission (e.g., via Wi-Fi) were introduced.
Network connectivity: Connectivity issues were investigated and optimised if possible.
Alert system: New alert messages and icons were introduced for real-time feedback.
Model integration: An analysis was conducted on how Googles Breezometer model maps onto the recorded trip geometries, which provided insights into air quality exposure along the routes.
2. DEVA trip recorder and DEV-D integration:
Data filtering and WHO guidelines: Data filtering functionalities were improved, incorporating references to WHO air quality guidelines.
User Interface (UI) and User Experience (UX): Usability testing led to improvements in the UI/UX, making the platform more user-friendly.
Graphical data visualisation: Visual representations of exposure data were enhanced, allowing for comparative analyses based on trip duration, distance, and absolute time.
Multinational testing: Field tests were conducted across various countries (UK, Austria, Spain, France, Germany, Italy, The Netherlands, Bulgaria, and Belgium).
Environmental condition testing: The system was evaluated under a wide range of environmental conditions, including urban and rural areas, indoor/outdoor settings, weather variations, and different times of day.
Interactive inhaled dose graphs: Exposure metrics in DEV-D were expanded to include interactive inhaled dose visualisations, allowing users to input their gender, age, and activity level for more personalised results.
Black Carbon (BC) sensor testing: A limited set of tests incorporated Black Carbon sensors without internal GPS functionalities toward the end of the COMPAIR project.
3. QADOS (formerly SODAQ) sensor:
Feedback and data on air pollution collected during testing were shared with QADOS to inform further product optimisations. This feedback encompassed the usability and technical characteristics of the mobile air quality sensors mounted on bikes using a holder and attached to backpacks with silicone sleeves and clips.
Outcome and future potential:
The culmination of the COMPAIR project is a well-calibrated and efficient toolset for monitoring air quality along travel routes using mobile sensors such as QADOS. Comprehensive user manuals have been developed for various configurations and use cases, enabling future adoption of the DEV-D and DEVA solutions for both policy-making and educational purposes.
With additional development and investment, new sensors could be integrated into the DEVA-DEV-D system. There is also significant potential for scaling the visualisation of modelled air quality data across Europe, which would allow air quality monitoring using just the DEV-D trip recorder, without the need for additional hardware.