The Responsible Organisation

Dublin City Council is the local government authority responsible for the city of Dublin, capital of Ireland. As a local government, it provides diverse and numerous services to citizens, including tourism development and promotion. In 2024, Dublin has been selected as the European Capital of Smart Tourism. 

The Tourism Unit within Dublin City Council specifically focuses on promoting and developing tourism in the city. They drive innovation and technology adoption to support the tourism industry in the city and are the implementors of the tool “A Day in Dublin”.

Smart Dublin is the unit within Dublin City Council dedicated to developing and implementing smart city solutions. It collaborates with various stakeholders, including technology providers, businesses, and citizens, to identify challenges and develop innovative solutions using technology. 

Data and Design is a startup specialising in data visualisation and storytelling. They transform complex data into engaging visuals and narratives that are easy to understand and interpret. They focus on communicating data-driven insights in a clear and compelling way. In this project, Data and Design's role to provide the AI-powered itinerary planner, through advanced data analytics, in an accessible and user-friendly format.

OpenAI is a leading AI research and development company. They are known for their work on advanced language models, including ChatGPT-4, which powers the "A Day in Dublin" itinerary planner.

For the preparation of this article, an interview has been conducted with the Dublin City Council and Data and Design, as well as with the participation of Barry Rogers, Head of Tourism for Dublin, and Rudy O'Reilly Meehan, the CEO of Data and Design.

The problem

On the one hand, like many popular tourist destinations, Dublin encounters the challenge of maximising the positive impacts of tourism flows to ensure equitable and sustainable growth, while also mitigating its negative effects. Some negative effects of tourism relate to overtourism, which occurs when a large number of visitors concentrate in popular areas, leading to overcrowding at attractions, strained infrastructure, and a negative impact on the environment. Overtourism can diminish the quality of the visitor experience and put pressure on local resources. This is further potentiated by traditional tourism recommendations, that often focus on well-known attractions, inadvertently contributing to overtourism. Guidebooks, websites, and travel agencies tend to highlight the "must-see" sights, leading visitors to the same places. This can result in long queues, crowded spaces, and a less enjoyable experience for tourists. 

Additionally, it might neglect the many hidden and unique experiences and places that a city like Dublin offers. Furthermore, generic recommendations fail to consider tourists’ individual preferences and interests. A one-size-fits-all approach to tourism doesn't consider the diverse needs and desires of visitors. For example, some tourists may be interested in history and culture, while others may prefer outdoor activities or food experiences. This way, visitors may miss out on opportunities that align with their specific interests and have, consequently, a less satisfying experience.

On the other hand, the City Council team has been experimenting with technology solutions and, particularly, AI-based tools across various city services for a long period now. In particular, following Barry Rogers, Dublin Head of Tourism, the Tourism Unit has developed a significant amount of knowledge and learning on tourism technologies, “largely for a few reasons, but mostly because Dublin is a great ecosystem for collaboration amongst [local startups] and large organisations and multinational firms like Google, MasterCard or in this case OpenAI”. Having been selected as the European Capital of Smart Tourism for 2024, the city has implemented a strategic smart tourism programme that comprises both community co-creation, innovation and open innovation processes, and technology adoption to improve the tourism sector. Therefore, there is an expectation that applying generative AI tools to tourism and fostering further collaboration presents the opportunity to experiment and learn how this type of tools can support specifically the tourism sector and industry, as Rogers shared in the interview.

The solution and its implementation

Dublin City Council, recognising the need to address the challenges described and aiming at experimenting Generative AI applications in the tourism sector, has partnered with Data and Design and OpenAI to create a touristic itinerary planner, called "A Day in Dublin" in 2024. This AI-powered tool aims to enhance the visitor experience by providing personalised itineraries tailored to individual interests and preferences.

At the core of this solution is the integration of GPT-4, OpenAI's language model, with advanced data analytics and visualisations provided by Data and Design. The tool combines user research to understand preferences through an automated chat function, data matching to align those preferences with local offerings, and AI-powered itinerary descriptions to create personalised suggestions. Specifically, through GPT-4, the tool poses questions to the users to understand their preferences and interests. The itinerary planner includes a user-friendly chat interface, allowing visitors to interact with the tool in a conversational manner. Users can input their interests, preferences, and desired pace of travel. This enables the solution to create a user profile, which orients the matching of the identified preferences and interests with the local offer of services, attractions, events, restaurant, and others, through advanced data analytics. The GPT-4 model is also employed to draft the narrative around the itinerary suggested, offering engaging descriptions for each recommended activity or attraction. In synthesis, while GPT-4 focuses on the interaction with the user, advanced data analytics concentrates on matching the identified profile with the available touristic offer from the dataset. This includes attractions, events, activities, restaurants, and transportation options. 

A distinctive aspect of this project was the collaborative approach that enabled its development and implementation. The City Council partnered with both a large global company, OpenAI, and a local startup, Data and Design. This reflects the "great ecosystem for collaboration" that Dublin is, as mentioned in the interview by Barry Rogers. Importantly, the team leveraged the existing experience of the Council with piloting solutions through an iterative and collaborative process, working with Smart Dublin, a dedicated piloting organisation. The familiarity with innovative procurement processes, as highlighted by Barry Rogers, is crucial. Since this project represented a small-scale prototype, an EU tender wasn't required and allowed for a certain level of flexibility. However, Rogers emphasises the importance of procurement processes for larger-scale adoption, for which the requirements of a full tender process are beneficial for de-risking and defining scope before investing resources.

Expected benefits

The itinerary planner is expected to bring numerous benefits to both visitors and the city of Dublin:

• The tool promises to enhance the visitors’ experiences of the city, by providing personalised itineraries. Tourists can receive recommendations that are tailored to their interests, expectations and needs.
• At a larger scale, the personalisation of itineraries promotes tourists to visit lesser-known attractions and distribute visitors more evenly across the city. This can help mitigate the negative impacts of overtourism.
• At the same time, the solution can promote positive spillovers of the tourism activity across a wider range of local businesses. The tool can support local businesses by recommending “hidden” or less known places, services, and experiences, driving traffic to establishments that might otherwise be overlooked. In the longer term, a wide adoption of this tool could potentially contribute to the growth of a more equitable local economy and promote a diverse touristic offer.

Main challenges 

While the tool offers significant potential, there are also challenges associated with its development, implementation and scale-up:

• Ensuring data quality is crucial for the tool's effectiveness. The local data used to generate recommendations must be accurate, up-to-date, and comprehensive. Maintaining data quality and relevance might require ongoing effort, resources and collaboration with various stakeholders in the future.
• Another challenge lies in mitigating bias in AI-powered recommendations. AI models data analytics tools can inherit biases from their training datasets, potentially leading to aleatory recommendations that can shape visitors’ behaviour favouring specific local businesses and services over others. Rudy O'Reilly Meehan, the CEO of Data and Design, states that “we need to have a conversation about the ethical implications of using AI in tourism". Having identified this, O'Reilly Meehan shares that a viable strategy to mitigate this risk can be incorporating a set of suggestions for each itinerary phase, instead of only one specific recommendation, giving visitors the opportunity to choose among a wider offer.
• Protecting user privacy and anonymity is also essential. As the solution is iterated and potentially scaled-up, Rogers points out that compliance with relevant regulation and the protection of users’ privacy is crucial. With this, it becomes important to explain transparently and openly how their data is collected and used.
• Finally, clear communication and education on the tool are necessary to address concerns about AI bias and build trust with internal teams, end users and other stakeholders. It's important to explain how the AI works, how it is being used to generate recommendations, and what steps are being taken to ensure fairness and accuracy.

Link: https://www.dublincity.ie/news/dublin-city-council-and-openai-announce-partnership-show-potential-ai-support-europes-tourism

Contact: tourism@dublincitycouncilculturecompany.ie