Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - CAMERA (Coordination and support Action for Mobility in Europe: Research and Assessment)

Teaser

\"At the turn of the century, Europe entered a new age, with the greatest demand for mobility in recorded history. While the global number of passengers in air travel has risen every year since the inception of commercial aviation, the arrival of the 21st century brought with...

Summary

\"At the turn of the century, Europe entered a new age, with the greatest demand for mobility in recorded history. While the global number of passengers in air travel has risen every year since the inception of commercial aviation, the arrival of the 21st century brought with it a significant increase in air travel. 2018 has seen 4.3 billion air passengers and, according to IATA, the projected number of passengers in the year 2035 is 7.2 billion (1). There are various reasons for this increasing demand in commercial aviation: low-cost airlines making aviation much more affordable, the expansion of economies and higher living standards especially in developing countries, the development of more fuel-efficient jets allowing airlines to provide more direct routes, greater urbanisation rate and better access to airports making aviation more accessible, etc.

European aviation is a crucial asset for economic growth and is a large wealth generator for the European Union. It provides vital transport links for the integration of Europe. The technologies and innovative concepts in aviation are catalysts for a plethora of other sectors, making its R&D initiatives one of the greatest return on investment. Therefore, it is important to identify gaps and challenges that threaten the sustainable development of the European air transport system. The EU designates significant funds for aviation research which should be optimised to properly address the needs of European citizens.

CAMERA aims at assessing the distribution of European funding across different research domains and the impact of those funds to achieve an \"\"integrated seamless, energy efficient, diffused intermodal system taking travellers and their baggage from door-to-door, safely, affordably, quickly, smoothly, seamlessly, predictably and without interruption\"\" for 2050 (Flightpath 2050). The analysis of the past and ongoing projects and their contribution to achieving these challenges, establishes the basis for future decisions and strategies at the EU level, to benefit its citizens and serve their mobility needs.
\"

Work performed

The first CAMERA reporting period, covering November 2017 to April 2019, focused on the definition, design, implementation and testing of the CAMERA assessment methodology. As a first step, the team collaborated in the development of the CAMERA Performance Framework to define the reference context for our assessment. The CAMERA Performance Framework defined the KPAs and KPIs to measure the success of European research in achieving the mobility challenges established in Flightpath 2050. In order to understand the complexity of the European air travel system and address the mobility challenges that the system is facing, CAMERA frames the whole door-to-door travel chain as the centre of its research. Referring to the vision outlined in Flightpath 2050, and the current state and future needs of aviation, five major thematic groups of challenges were identified:

- creating an individualised and seamless mobility system for everyone;
- improving the overall performance of the mobility system;
- improving the resilience and re-configuration of the mobility system;
- providing safe and efficient air traffic management services;
- designing and implementing an integrated, intermodal transport system.

A set of key performance indicators (KPIs) for each of the five layers were defined in the PF to assess the contribution of European research to address these challenges.

Once the contextual framework was defined, CAMERA aimed at identifying those relevant projects whose scope could create an impact across the European mobility system, addressing the challenges previously identified. The full list of FP7 and H2020 projects, with a total of just over 40 000 projects\' data, was cleaned, prepared and analysed using natural language processing techniques to identify those projects which were contributing to progress beyond the state of the art in the area of mobility. A total of 158 projects were selected by the unsupervised AI algorithm. These projects were taken forward for further analysis using semi-supervised clustering techniques from machine learning. The result of the clustering algorithm provided a probability distribution per project of the likelihood of addressing each of the 9 topics automatically found by the model, and \'self-defined\' as a set of key words. A second analysis took the five layers defined in the PF as a reference for the clustering. For this exercise, additional textual data (reports, research papers), describing the layers was used to train the text-mining models to understand the scope of each of the layers.

The descriptive analysis of the 158 selected projects showed their geographical distribution across Europe and historical evolution from 2007 to 2018.

From a content-wise perspective, the topics (word clouds) automatically detected by the algorithm were assigned titles by the CAMERA team to simplify their reference:

- Air transport
- Ground transportation
- Intelligent transport
- Freight transport
- New concepts in transport
- Environmental impact
- Socio-economics
- Intermodality
- Urban mobility

A fairly even balance across the topics was observed, although with some dominant ones. In particular, \'intelligent transport systems\' is the area with the highest number of projects, followed by \'intermodality\', \'ground transportation\' and \'air transport\'.

Final results

An innovative artificial intelligence-based methodology has been designed over CORDIS data, supported by expert-based assessment and validation. CAMERA uses CORDIS data provided by the EU Open Data Portal as the main source of information. CORDIS is the European Commission\'s primary public repository and portal to disseminate information on all EU-funded research projects and their results. Currently, it covers just over 40 000 projects from the FP7 and H2020 programmes. This repository contains many unclassified and unstructured texts. Traditional analysis techniques are not able to effectively handle this kind of unstructured data. In order to implement a system that can effectively extract crucial semantics from these texts and classify projects into various categories, natural language processing (NLP) techniques have been applied.

In CAMERA, we experimented with a semi-supervised text classification technique based on a latent Dirichlet allocation algorithm that has the ability to detect underlying patterns in analysed textual data, and based on those patterns, to group the text documents into separate categories, without needing any input from the human actuator. The algorithm was used to filter mobility relevant projects from the full set of (over 40 000) projects and automatically classify them into nine different groups. Furthermore, based on the Performance Framework produced by the experts of the CAMERA consortium, and an additional set of relevant, human-labelled documents, the abstracts and summaries of the extracted projects were analysed semantically. Based on algorithmically-calculated semantic similarity, the mapping between the projects and CAMERA-defined mobility layers, reflecting the high-level objectives of Flightpath 2050 and a number of other strategic documents, was performed. For this mapping, a number of algorithms was experimented with and implemented.

Website & more info

More info: https://h2020camera.eu.