Transport

Introduction 

The transport domain has been a driving influence in the open data movement since the beginning. Its importance could be attributed to the significant and direct impact that transportation has in our daily lives, influencing how much time we spend getting to school or work and determining where we choose to live. So it should not be a surprise that technological changes that influence how we inform ourselves and interact with our means of getting from point A to B are often under the spotlight.

The 2019 State of Open Data chapter on Transport took stock of the development of data-driven transport tools and highlighted tensions between centralized route planning services and distributed, open data-driven approaches to transport data. This update considers what has happened since the original publication and outlines the challenges that remain to fulfilling the true potential of open transport data.

The last five years have not seen large technological changes in how open transport data is published. However, there has been significant progress from an administrative standpoint as public authorities have further recognized the value of unlocking data. Open transport data reuse, with AI-based technologies driving innovation, has also seen significant progress. One example of this is the use of data for tackling climate change and urban sustainability. 

Well-known challenges, such as data interoperability, remain at the core of technology development. The discussion around data privacy, brought forth by policies such as GDPR in Europe, has also introduced new technical challenges for the open transport data ecosystem.

Nevertheless, the future seems bright for open data in the transport domain with an increasing amount of new data sources being disclosed that are powering the growth of innovative applications to tackle real world problems. In certain cases, public authorities who realize the importance of open transport data are making it a core asset for their decision-making processes. 

The Popularity of Open Transport Data

Judging by the number of users, public transport continues to be the most successful use case for open transport data. For example, Google Maps reports having over 1 billion users every month.¹ This success is mainly due to the ubiquitousness of the General Transit Feed Specification (GTFS), a data specification that allows public transit data to be consumed by a wide variety of software applications. One of the main reasons for the wide adoption of GTFS may be its open and democratic governance process² that allows anyone to propose specification changes, discuss them with the community, and vote on their adoption as part of the standard. 

Following in GTFS’ footsteps, 2014 saw the introduction of the General Bikeshare Feed Specification (GBFS). Driven by the increased popularity of shared mobility services, such as bicycles, scooters, and mopeds, it follows a similar open governance process as GTFS,³ and its adoption continues to increase with more than 700 known data publishers to date.⁴

However, open transport data is not restricted to public mobility. There are more and more open data sources focusing on transport infrastructure, a core aspect of the transport domain. Open infrastructure-related data sources can be found on official public data portals at the continental (e.g., Europe), national (e.g., US), regional (e.g., New South Wales, AU), or local level (e.g., São Paulo, BR). 

One of the most prominent open data sources is OpenStreetMap (OSM), a flagship of the open data movement. OSM contains information about various topics, such as buildings, points of interest, and natural formations, but one of its main focus areas is road and railway networks. An article published by Barrington-Leigh et al.⁵ estimated that OSM data covers about 80% of the world’s road networks. In many countries, OSM constitutes the only open data source about their transport infrastructure. OSM’s quality continues to improve as the number of contributors grows. 

Measuring Impact

In the past, the economic impact of open transport data was assessed by looking into the number and value of applications created from the data and also by how many people were reached by those applications. Nowadays, the impact of open transport data has reached the point of influencing policy and even changing the way public authorities manage their data. 

For example, in Europe, open transport data is considered one of five highly impactful data domains.⁶ Its importance led to the creation of the Intelligent Transport Systems directive for National Access Points (NAPs).⁷ This directive requires every member state of the EU to create a single access point where relevant and open transport data sources are made available to facilitate the creation of multimodal and cross-border transport services. A NAP list and interactive visualization is made available online.⁸ 

The "Pin je punt" initiative operated by the Flemish region of Belgium’s tourism agency is another powerful example of open transport data impact.⁹ Toerisme Vlaanderen encourages citizens to use a web application to map things, such as transport-related infrastructure (bike pumps, bike sharing places, or e-bike charging spots), by registering their geographical location, taking photos of them, and recording other characteristics. The data is fed directly into OSM and is carefully reviewed and maintained by public officers. 

Such examples demonstrate how the evolution of open and crowd-sourced transport data has reached a level of maturity and trustworthiness that allows public authorities to depend on it and further invest in its development.

Trends in Open Transport Data

The last decade has seen impressive advances in artificial intelligence (AI) technologies with machine learning being at the centre of innovation. This technological development has also fuelled the creation and improvement of many applications and services in open transport data. 

In 2019, Niedstat et al.¹⁰ presented a report to the European Commission about the current and future developments of AI in the transport domain. The report analyzed the applications of AI in different transportation sectors at the time. Summaries of its findings are listed below:

• Road transport: AI technologies for the road transport sector were focused on developing self-driving cars or trucks.¹¹ Such vehicles rely on a variety of sensors to be aware of their surroundings, requiring open data such as road network topology, traffic signs, road disruptions, and parking locations. Other applications can be found in road traffic management with e.g., AI algorithms adjusting traffic light intervals based on traffic flow¹² or real-time bike infrastructure monitoring.¹³ Along the same lines, Google announced the use of machine learning techniques to predict bus service delays drawing on historical records, road infrastructure, and traffic congestion.¹⁴ 

• Air transport: Optimization and automation are the main goals of AI technologies applied to the air transportation sector, where open data has always had an important role.¹⁵ Multiple research projects focus on issues like air traffic management and predictability or improvement of passenger flows.¹⁶

• Rail transport: AI technologies applied to the railway sector mainly aim to increase the automation of train operations, focusing on self-driving trains but also traffic management and maintenance planning. Open infrastructure data is crucial to that end, especially in cross-border scenarios. In 2021, the European Agency for Railways (ERA) published the ERA Knowledge Graph as Linked Open Data,¹⁷ which provides EU-wide railway infrastructure information for automating vehicle compatibility assessments.¹⁸ OSM is also a rich data source for railway infrastructure, which can be visualized with the OpenRailwayMap application.¹⁹

• Maritime and river transport: One of the most important data resources in the marine transport sector comes from the Automatic Identification System (AIS), a tracking system that draws information from transceivers placed on ships. Machine learning applications aim to use this data for predicting vessel arrival times and also to prevent collisions. An example of open data is found on the United States’s National Oceanic and Atmospheric Administration’s website, which publishes historical AIS data.²⁰ Applications and services have been created around AIS data, such as the MarineTraffic Web application²¹ that allows visualization and tracking in real-time of vessels around the world. 

Open transport data has also proven crucial in recent efforts to tackle issues such as climate change. Initiatives, such as the Connecting Europe Facilities (CEF) GreenMov project launched in September 2021,²² aim to develop open transport data powered services that promote greener transportation and help reduce CO2 emissions. Along the same line, Yadav et al.²³ presented a study of the role of open data for sustainable mobility in nine major cities around the world. They characterized and compared different available open data sources in each city and discussed how they showed signs of driving innovation for sustainable growth.

Mobility as a Service (MaaS) continues to be discussed as one of the cornerstones of urban sustainable development. Its goal is to provide travellers with a single interface to plan, book, and pay for integrated and multimodal mobility alternatives for a seamless travel experience. Open transport data is crucial in such a scenario as it often requires the integration of multiple and independent mobility providers. 

A 2021 report from the World Bank²⁴ highlights the importance of MaaS for cities as an organizing framework for understanding and shaping how multiple mobility alternatives can work together to achieve sustainability and development goals. Yet, fully reaching MaaS ideals requires a multidisciplinary effort led by public authorities and not only focused on technological challenges. Government action through policies and regulation is fundamental to creating a framework for MaaS to be developed under fair conditions and to meet the mobility needs of all sectors of the population - not only those of higher income and technology-enabled travellers. 

Remaining Challenges

Visions of the future where transport data is used for the good of all, like those espoused by MaaS or by Data Space-based scenarios,²⁵ recognize data interoperability across systems and organizational boundaries as a fundamental requirement. But how can this be done when there are numerous means of transport and many different stakeholders? 

Open data can be seen as one of the first steps to achieve interoperability, but is certainly not the last. The MaaS Alliance,²⁶ a public-private partnership working to establish the foundations for building a common approach to MaaS, identifies a set of challenges that need to be addressed to fully reach data interoperability. These are:²⁷

• Identifying existing standards and data models;
• Agreeing on what to use at the current state of knowledge;
• Studying the possible or necessary evolutions;
• Convincing different stakeholders to use these solutions.

Open standards constitute a valuable step towards interoperability. However, aligning and agreeing on common and unambiguous understandings of concepts from such standards (also known as semantic interoperability) is still a difficult task that can only be achieved through collective effort.

Another important challenge that data publishers need to consider when disclosing data is personal data protection and privacy. Recent privacy regulations like GDPR in Europe establish a legal framework for the collection, processing, and disclosure of personal information to protect against misuse, unlawful access, alteration, or loss of such information. Initiatives such as MaaS or Europe’s Intelligent Transport Systems directive²⁸ require data to be open and exchanged across organizational boundaries, but they do not specify rules related to data protection and privacy beyond mentioning the need to comply with applicable legal frameworks.²⁹ The risks that arise when opening transport data without proper access control include:³⁰

• Identification of individuals;
• Tracking of individuals through the transport network;
• Breach of privacy laws;
• Privacy attacks;
• Reconstructing past trips of specific individuals; and
• General safety of city infrastructure, vehicles, property, and public transport personnel.

Therefore, it is important that data publishers make sure that their open data sources are properly anonymized and that they have thorough revision procedures in place. Recent architectural specifications for decentralized data exchange (e.g., GAIA-X,³¹ Solid,³² and IDSA³³) foresee the definition of policy-based access control mechanisms that aim to ensure compliance with data protection and privacy regulations.

Conclusion

The last five years have continued to bear witness to the value that open data brings to the transport domain. Open transport data is a fundamental asset in a number of new applications and services driven by the development of AI technologies, which are being applied on the different transport sectors, including road, air, rail, and maritime. It also plays a crucial role in initiatives like MaaS for tackling global issues, such as sustainable urban development and climate change, by helping to make the shift toward greener and more efficient mobility alternatives.

Still, reaching the full potential of open transport data requires continued technological as well as administrative effort. Technologies need to be developed for facilitating data interoperability. This, in turn, requires that collaborative legal, economical, and technical frameworks are put in place to allow for multiple organizations to interact and efficiently exchange their data while complying with data protection and privacy regulations. 

Despite these challenges, the future looks bright for open data in the transport sector. We can only wonder about what exciting new developments will come in decades to come with the certainty that they will continue to change our lives.