VRUs and micro-mobility safety services V1

SOTERIA will develop multiple information services that will assist Vulnerable Road Users (VRUs) 
with minimising errors in their decision-making, avoid exposure to hazardous situations and 
receive proactive notifications for potential hazards. This deliverable presents the first (Alpha) 
release of the sensors developed in T3.1 | Micro-vehicles active safety systems and the services
developed in T3.2 | VRUs and micro-mobility awareness safety services, T3.3 | Safety-oriented 
proactivity and nudge engine and T3.4 | Safe and clean routing (SARA-VRU). It describes the 
developed two-wheeler mountable sensor kit as well as the detailed design of safe travel services 
and the corresponding software implementations. These safe travel services for supporting VRUs 
in their day-to-day traveling include:
▪ micro-mobility hazards mapping
▪ situational aware speed advice
▪ real-time protective equipment assessment
▪ a nudge engine that recommends safe behaviours, and
▪ a safe and clean routing engine that incorporates intelligence related to areas of the 
infrastructure that need to be avoided.
An initial prototype of the SOTERIA sensor kit has been developed and integrates several sensors 
and micro-controllers (i.e. a Raspberry Pi 4 and Arduino Portenta H7 board) on micro-vehicles, 
enabling the collection of heterogeneous contextual, vehicle handling, kinematics and 
environmental data. This sensor kit includes an accelerometer, a gyroscope, an environmental 
sensor, a pulse sensor and five ultrasonic sensors. Additional sensor kits support the installation 
on both public and private bicycles, motorbikes, e-scooters etc. Collected data are geolocated to 
trace possible pollution hotspots and they are periodically transmitted, using a wi-fi network, to 
the SOTERIA Safe Mobility Data Space (SMDS). ONSEE's smart bicycle camera (REBO) ensures 
every near miss and road issue is captured at the push of a button on the bicycle’s handlebar (it 
saves recordings of events 30 seconds before and after they happen). The REBO platform will be 
synchronised with the SOTERIA system through a secure RESTful Application Programming 
Interface (API).
Exploiting the real time micromobility data collected by the sensor kit, the micro-mobility hazards 
mapper will identify the locations, types and severities of all potential accidents and hazards. 
However, the initial version of the service provides VRUs with a map of hazard hotspots, since it
has been developed for detecting different types of hazards. There is not actual sensor data 
available at this stage of the project and therefore, a comprehensive collection of open data has 
been used to develop and test the potential functionalities of the micromobility hazards mapper. 
A Machine Learning (ML) algorithm is proposed, which has been applied to the full set of 
acceleration data (components along three axes) and the distance data from nearby vehicles to 
detect micro-vehicle–vehicle encounters. Additionally, the methodological approach, 
specifications and technical design of the speed advisory module and the protective equipment 
assessment module are presented and demonstrated.
The fundamental aim of nudge engine is to deliver nudges, which in essence are strategically 
designed notifications and suggestions, to promote safe behaviours among road users and alert 
them of imminent dangers. It utilises a comprehensive nudge repository (i.e. message pools 
stored in a MongoDB database) and a decision-making framework that filters and selects the most 
appropriate nudge messages for the users at every point (i.e. pre-trip, on-trip, and post-trip).
A safe and clean routing algorithm has been developed, considering the outcomes of accident 
prediction models to define safety per section and weights for routes by summing up risk values 
for all used sections of a route. Safe routes can be calculated using route optimisation engine like 
OpenTripPlanner or Open-Source Routing Machine (OSRM). Additionally, the SOTERIA journey 
planner is designed to provide a variety of travel options involving multiple modes of 
transportation, while also incorporating safety and environmental considerations for VRUs and 
public transport users.