Content

AI in dialogue: improving subjective and objective safety in bicycle and pedestrian traffic - Civic Innovation Platform

Ai in dialogue

Improving subjective and objective safety in bicycle and pedestrian traffic

The company FixMyCity GmbH,HafenCity University Hamburg (HCU), the association CorrelAid e.V. and the city of Aachen intend to develop an AI application that can be used by public administration when planning safe infrastructure, with the aim of contributing to greater bicycle and pedestrian traffic safety while teaching the users how to use AI in a fun and intuitive way. Its aim is also to offer all users easy and tangible access to AI in the area of bicycle and pedestrian traffic.

Why are you a strong team? 

We are an interdisciplinary team of transport experts, academics, data scientists, software engineers and representatives from local authorities. Pedestrian and cyclist safety is something very close to us, which is why we are taking a stand for increased safety to and from work and school with our project. With our concentration of expertise, we will ensure that our AI solution is implemented the right way for its audiences, with results that add value.

Explain your idea in three sentences. 

We want to identify danger points on the journey to and from work or school so that they can be considered in transport planning, potentially reducing the number of accidents. The purpose is to benefit individual welfare and reduce costs on the economy. AI is used in the form of the personal assistant ‘Kioma’, which helps users record these danger points using a web app and processes the data in a structured way for local authorities.

What makes your idea special?

The AI is intended to come across as a trustworthy and accessible personality. We are making sure that the design is appropriate for the audience, given that it addresses schoolchildren. Important information is collected and processed in a standardised manner using an intelligent dialogue system. Insufficient volumes of data are improved by the AI by automatically adding the opinions of other users. The solution developed will then be transferable to further audiences.

What are the next steps? 

We are first developing concepts for visualisation and interaction with the web app that are tailored to different age groups. We are focusing on preschool- and school-aged children between 3 and 16. Based on this, we will test various designs while involving parents and schoolchildren. Together with representatives from local authorities, we will define a concept for the data infrastructure in which the collected information will be stored for further use.