At a Glance

Project acronym: REDUCTION

Project type: Specific Targeted Research Project (STREP)

Programme: 7th EU Framework Programme

Project coordinator:
Lars Schmidt-Thieme
Inf. Systems and Machine Learning Lab
University of Hildesheim
Tel: +49 5121 883 851

Project partners:
Stiftung Universität Hildesheim (DE),
University of Thessaly (GR),
University (DK),
Aarhus University (DK),

Delphi Delco electronics GmbH (DE),
Trinité Automatisering B.V. (NL),
Nordjyllands Trafikselskab-Bektra (DK),
TrainOSE S.A. (GR),
CTL Cyprus
Transport Logistics Limited (CY)

Start date: Sep 1, 2011
End date: Aug 31, 2014


REDUCTION follows an interdisciplinary approach bringing together expertise from several communities, such as: Data collection and management, which will develop the mechanisms for handling the large volumes of streaming data; Data Mining/Machine Learning, which will be responsible for developing decentralized algorithms for predictive analytics; Wireless Networks, for developing protocols for effective vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications; Vehicle electronics, which will develop the on board computing and sensor devices; Transportation engineering, which will be involved in the evaluation of the system and the data generated during the field-trials.

REDUCTION will develop powerful methodologies for providing predictive analytics based on advanced data mining technology, which will reveal patterns and useful information for meeting the objectives of fleet management, such as decision making for driver-adaption, ecorouting, and thus CO2 emissions control and improved fuel economy. Web-based access will be provided to clients of the provided service, in order to provide explanatory reports, based on graphical visualizations, about the discovered knowledge and for providing rapid alerting information that enables monitoring the performance of the fleets. Finally, the REDUCTION platform will provide appropriate interfaces to intelligent mobile devices (e.g., smartphones, netbooks) in order to enable public user - e.g., passenger - applications leverage the information residing in the platform.


Work Package Description


Work Package 1 (WP1):

Onboard Technology and Wireless Communication


Work Package 1 deals with basic communication infrastructure and wireless communication. Its objective is to develop the on-board technology taking also into account the requirement for supporting multi-modal fleets.



Work Package 2 (WP2):

Predictive Analytic Models for Energy-Efficient Driving and Driver-Behavior Adaptation


The main objective of WP2 is to develop novel algorithms for creating predictive analytics models that will operate in the decentralized environment of REDUCTION. The proposed prediction models will enable the generation of knowledge for supporting driver-behavior adaptation in order to educate drivers about ways of energy-efficient driving. Therefore, this work package has also the objective of developing distributed data mining algorithms that will be designed to run onboard and exchange computation and data with other vehicles and with the centralized infrastructure.


Work Package 3 (WP3):

Data Management for Environment Aware Routing and Geo-Locational Analysis Application


The objective of WP3 is to design and develop a software prototype that can convert vehicle-related data, primarily GPS data, to metrics that capture environmental impact. The prototype must handle very large volumes of data from different types of vehicles and must efficiently compute the multi-modal eco-routes in both real-time and offline modes. In addition, the prototype must be able to report on the temporal evolution of eco-routes, e.g., due to a variety of changes in the transportation infrastructure and its use.


Work Package 4 (WP4):

System Design & Integration


The main objective of WP4 is to have a real-time publish-subscribe distributed middleware with a generic functionality. Components may subscribe for information, unsubscribe, publish information, and notify that they are interested in some kind of information. The Event Handler receives all these events through an interface, and can also notify components when the information is of relevance for them. The publish-subscribe communication mechanism will support an asynchronous (non-blocking), many-to-many communication between components in the network.

Work Package 5 (WP5):

Case Studies for assessing Energy-Efficiency and CO2 Reduction


The goal of this task is to twofold; firstly, to confirm that the architecture of the REDUCTION system is generic enough to encompass diverse “application” scenarios, and secondly, to provide useful input to the partners for any omissions concerning the operational part of the system, that might have got unnoticed, or to develop more advanced features for the system. Therefore, the existence of several field trials is mandatory, multi-modal and traditional as well.

A Snapshot of application from TRAINOSE field trial

Work Package 6 (WP6):

Dissemination, Exploitation, Standards


The results of REDUCTION are made publicly available through peer-reviewed publications, conference presentations, press releases, web pages and brochures. Generated intellectual property is carefully protected, e.g., patent filling. An exploitation and dissemination plan is specified to maximize the outcome and benefit of the project for individual partners. Partners will search for and use existing fleet-management standards used in the EU. Important contributions to these standards are made where applicable.

Work Package 7 (WP7):

Project management


This WP will ensure the effectual and timely achievement of goals within the project in the most cost-effective manner