hyBrid bUs geoSpatial Data and Network Analysis

SCHEME: Industrial Fellowships

CALL: 2018

DOMAIN: IS - Information and Communication Technologies

FIRST NAME: Sune Steinbjoern

LAST NAME: Nielsen


INDUSTRY / PPP PARTNER: Luxembourg Centre for Systems Biomedicine (LCSB), University of Luxembourg

HOST INSTITUTION: E-Bus Compentence Center

KEYWORDS: electromobility, bus fleet management, big data, machine learning, visual analytics, interdisciplinarity




Submitted Abstract

Electrification is a disruptive innovation in the bus industry, rapidly changing the technology scene. Electrified buses (e-buses) are complex systems using several constantly evolving battery and charging concepts. Thus, the question of selecting an optimal e-bus type (hybrid, plug-in hybrid or electric) and its configuration (battery size, charging type, etc.) is very difficult due to the multi-objective nature of the problem. E-bus system performance not only depends on the bus itself but also on its operational characteristics, defined by the route topography, weather and traffic conditions. Rapid technological changes and system complexity cause confusion in the entire public transport industry, which is not used to changes at such a pace. Consequently it hampers the adoption of e-buses. The goal of this project is to design and validate new methods allowing to understand, predict and improve performance of e-buses in reference to specific operational conditions. These methods will replace currently used rules of thumb based on average operational characteristics. Novel, data-driven visual analytics methods, currently employed in the field of biomedicine and clinical research, will be used on the new type of event-based data, generated by e-buses during real-time operations. This will enable experts to explore complex datasets and gain deep insight into their content and structure, allowing to determine causes of inefficiencies and to proactively react when the conditions occur. The combination of large amounts of collected data with precise geocomputational city models, applied visual analytics and incorporation of machine learning methods will be used to predict the impact of new operation strategies. This will allow for optimal organization of existing fleet operations, improvement of electrical-to-fuel drive strategies, and projection to new use-case scenarios in other cities.

This site uses cookies. By continuing to use this site, you agree to the use of cookies for analytics purposes. Find out more in our Privacy Statement