Road congestion and the difficulty of finding parking spaces impede people’s mobility and slow down the delivery of goods and services. Large amounts of data from diverse sources could help provide solutions for better planning and management of urban traffic. However, the data lacks harmonisation and is often ‘noisy’ or out of date.
The EU-funded QROWD project provided innovative solutions for local government and transport businesses to improve mobility, speed up deliveries and make travelling more efficient, safer and greener. This was achieved through a new approach to integrating different sources of data, including geographic, transport, meteorological and real-time data about individuals, crowds, infrastructure and public transportation.
‘QROWD delivered a platform to design transport and mobility services in cities and to engage with citizens, residents and visitors, and city planners who collect the data on those services,’ says project coordinator Elena Simperl, professor of computer science at the University of Southampton, in the UK.
The platform was piloted in the Italian city of Trento. ‘We have a much better understanding of mobility patterns in Trento to improve traffic management, and better data for making traffic predictions,’ Simperl says. ‘We are now looking at how the platform can be packaged for use in other cities.’
Learning from big data
A key innovation was bringing together artificial intelligence (AI) – machine learning – and crowdsourcing to develop transportation solutions. The QROWD platform incorporates a new hybrid computer system using algorithms combined with real-time human feedback. To achieve this, the project team overcame key technical and non-technical challenges.
The main technical challenge concerned how to apply machine learning to identify and analyse big data – the huge amount of relevant, heterogenous and often complex data from different sectors in diverse formats. Integrating such data lies beyond the capabilities of traditional data-processing software, while for some transportation challenges additional data must be collected.
The project team developed new tools for data collection, including a mobile app that revealed people’s mobility patterns – such as the daily commute to work – and to help the city keep its transport data up to date, by confirming for instance where disability parking spots are located. These tools were developed through industrial partners: satnav producer TomTom, and Atos, AI4BD and Inmark Europa, which specialise in big data, IT and AI.
Citizens help provide solutions
The main non-technical challenge was how to develop participatory approaches with citizens to co-design transport-related services, using data collection at a time when the use of personal data is a growing concern. This led to the development of a crowdsourcing platform that reflects ethical concerns.
Crowdsourced data enabled the project team to look beyond IT to consider behavioural, economical and psychological aspects of mobility, putting a citizen-driven approach at the heart of smart city development.
Citizen involvement enhanced smart city mapping, leading to a virtual city explorer tool. The benefits include reduced travel times and easier location of parking spaces or bicycle racks.
This leads not only to more efficient transport but also to a better urban environment. Traffic pollution and greenhouse gas emissions arising from traffic congestion can be reduced, for instance, ultimately enhancing the quality of life in European cities.
‘The service will be used by city planners, giving them a way to think about how they engage with citizens to design better transport policies,’ says Simperl. ‘There has been interest from several cities.’ Particularly interested are European cities involved in the Open & Agile Smart Cities group which shares the focus on implementing open platforms with citizen engagement.