# A New Take on Protecting Cyclists in Smart Cities

## TL;DR

This paper proposes a novel algorithm using geofences and hybrid vehicle actuation to dynamically reduce pollution exposure for cyclists in smart cities, validated through simulation and real-world testing.

## Contribution

It introduces a new geofence-based control algorithm for hybrid vehicles to lower emissions around cyclists, enhancing urban pollution protection.

## Key findings

- Algorithm effectively reduces emissions within geofences.
- System successfully tested in simulation and real vehicle environments.
- Potential to improve cyclist safety in polluted urban areas.

## Abstract

Pollution in urban centres is becoming a major societal problem. While pollution is a concern for all urban dwellers, cyclists are one of the most exposed groups due to their proximity to vehicle tailpipes. Consequently, new solutions are required to help protect citizens, especially cyclists, from the harmful effects of exhaust-gas emissions. In this context, hybrid vehicles (HVs) offer new actuation possibilities that can be exploited in this direction. More specifically, such vehicles when working together as a group, have the ability to dynamically lower the emissions in a given area, thus benefiting citizens, whilst still giving the vehicle owner the flexibility of using an Internal Combustion Engine (ICE). This paper aims to develop an algorithm, that can be deployed in such vehicles, whereby geofences (virtual geographic boundaries) are used to specify areas of low pollution around cyclists. The emissions level inside the geofence is controlled via a coin tossing algorithm to switch the HV motor into, and out of, electric mode, in a manner that is in some sense optimal. The optimality criterion is based on how polluting vehicles inside the geofence are, and the expected density of cyclists near each vehicle. The algorithm is triggered once a vehicle detects a cyclist. Implementations are presented, both in simulation, and in a real vehicle, and the system is tested using a Hardware-In-the-Loop (HIL) platform (video provided).