Three levers to unlock artificial intelligence benefits in urban mobility

With its ability to interpret data, assist judgments, and generate insights that are beyond human manual processing capabilities, artificial intelligence (AI) is profoundly changing urban mobility systems.

Promising AI applications focusing on end users’ journeys are at the centre of the recent EIT Urban Mobility report on AI[1], which explores the use of AI in fields such as Mobility as a Service and Demand Responsive Transport. One of the report’s use cases features HaCon and Siemens Mobility, two companies that work with machine learning to predict the number of passengers in public transport, providing travellers with estimates of how crowded public transport will be. The solution has been implemented by the Rhein-Main-Verkehrsverbund (RMV) in and around the city of Frankfurt in the middle of the COVID-19 pandemic to make social distancing easier for public transport users.

This is a testament to the fact that AI is already playing a central role in supporting the development of user-centric, citizen-oriented urban mobility.

Similarly, the AI TraWell project[2], co-funded by EIT Urban Mobility and led by a consortium including Fraunhofer and University College London, is using AI to create an app that optimises multimodal routes to fit travellers’ preferences and wellbeing. In this case, AI helps provide predictive information about the different transport modes a user considers, so each journey can be adapted to one’s preferences. Additional AI applications in mobility include traffic and logistics forecasts, on-demand bus services, vehicle tracking on transport networks, or self-driving car technologies such as automated valet parking, which is a currently widespread application for AI in cars.

To further accelerate the development and deployment of AI in the mobility sector, business leaders and policymakers across Europe need to pay special attention to three priority areas: investments in AI; data infrastructure & skills; ethical & trustworthy AI.

1. Increase AI investments for Europe to catch up in global race

With its potential to increase efficiency and productivity, the momentum around artificial intelligence has been growing over recent years, in sectors as diverse as information and communication, wholesale and retail trade, and manufacturing. In terms of yearly investments, the US and China are clearly dominating the international stage: of the €25 billion total investment worldwide in blockchain and AI technologies each year, the US and China account for more than 80%, while the EU’s share only amounts to 7%, or about €1.75 billion. This echoes findings from the EIT Urban Mobility report on AI, according to which one of the biggest barriers to AI solution deployment arebudget constraints (for a third of the survey respondents).

Stressing this investment discrepancy between Europe, the US, and China, a recent report on AI and blockchain by the European Investment Bank (EIB) highlights a gap of up to €10 billion that is “holding back development and deployment of artificial intelligence and blockchain technologies in the EU.” The launch of a new EU funding instrument in March 2021 of up to €150 million should help close part of this gap, by focusing on AI early and growth companies in Europe.

2. Improve data infrastructure and skills to reap the full benefits of AI

According to the results of the industry survey[3] of the EIT Urban Mobility study, the provision and use of AI services in Europe is still mainly focused on the business-to-business (B2B) sector: of the AI companies surveyed in the EIT Urban Mobility report, 76% operate within the B2B space. While all companies use or provide AI in part of their daily operations, more than 60% of survey respondents have yet to feel that AI solution deployment actually leads to revenue generation, and 54% of respondents stated that AI solutions had not been impacting costs. This reflects some important barriers that are still hindering the full use of AI’s potential. In this regard, the lack of technical feasibility (stressed by 60% of respondents) and the poor data availability and quality (50%) are the most important non-financial bottlenecks.

The positive impact of AI on revenue increase and cost reduction still needs to be felt in some key business areas 

In parallel, the governance of both AI providers and adopters needs to be improved in line with the new challenges and opportunities arising from a more widespread use of artificial intelligence in businesses: according to the experts surveyed, the three main risk areas where improvements are needed to better fit AI in corporate activities are cybersecurity (73%), compliance (55%), and personal privacy (55%). To face these risks, European businesses need competent staff equipped with the necessary AI knowledge and skills to make the best use of AI and reap its full benefits.

3. Implement a clear framework for ethical, trustworthy AI

Despite being already omnipresent in our lives, AI is not regulated under a set of globally agreed standards. However, in the global race for competitiveness in AI, the EU stands out as a leader in ethical AI, thanks to an encompassing data strategy. In April 2021, the European Commission proposed a harmonization of rules for AI to establish a framework protecting people’s security and fundamental rights. This will be an essential milestone toward the development of a single market for lawful and safe AI, while preventing market fragmentation. Importantly, it will contribute to reinforce trust in the technology and provide the regulatory certainty cutting-edge innovation needs to thrive.

Looking ahead, enhanced governance and more effective enforcement of existing regulations applicable to AI systems will accelerate the development and deployment of AI in Europe, with positive impact on the accuracy and uptake of user-centric urban mobility applications.

The report can be accessed here

Background information

[1] Report based on a survey of more than 60 AI experts in the fields of urban mobility, climate, manufacturing, and health – both AI technology providers and adopters.

[2] AI TraWell is a consortium project of Achmea, Eindhoven University of Technology, Fraunhofer Society, Gehl Architects, ISBAK, Münchner Verkehrs- und Tarifverbund, Open & Agile Smart Cities, TomTom, University College London, with the participation of the cities of Copenhagen, Istanbul, Lublin, and Munich.

[3] The survey sample of 64 replies includes companies and institutes affiliated with EIT Climate, Manufacturing, Urban Mobility, and Healthcare. The survey covers both data management for AI development as well as AI business model and applications in Europe. Respondents are mostly AI solution providers (68%), followed by AI technology adopters (28%), and AI technology investors (4%).