Advances in location based services

1.

The research field of Location Based Services has a long tradition. The Location Based Services conference series has been organised since 2002. It has become one of the most important scientific events on the topic LBS. Since 2016, the conference is the main annual conference for the International Cartographic Associations (ICA) – Commission on Location Based Services. Previous conferences were organised in Vienna (2002, 2004, 2005), Hong Kong (2007), Salzburg (2008), Nottingham (2009), Guangzhou (2010), Vienna (2011) and Shanghai (2013), Vienna (2002, 2004, 2005, 2011, 2014, 2016, 2019), in Munich/Augsburg (2012, 2015, 2022), Zurich (2018) and Glasgow (2021). For Location Based Services 2022 in Munich/Augsburg, 12 full papers and 21 extended abstracts were accepted for the conference after a peer-review process. Following the successful conference, authors of the selected papers were invited to submit an elaborated version of their papers for this special issue of the Journal of Location Based Services. Seven papers were accepted for the special issue.

Therefore, this special issue provides a collection of peer-reviewed, up-to-date research articles on the topic of location-based services. Because progress at LBS is broad, the articles document research activities from a variety of fields. To illustrate the broad range of topics covered in the conference series, and in this special issue in particular, we have computed illustrative images for each of the papers. All of these figures were created with the Bing Image Creator, which uses DallE artificial intelligence algorithms. The images are created in July and September 2023 using the underlying prompts. The results are documented below.

The paper by Xu Feng, Khuong An Nguyen, and Zhiyuan Luo-‘WiFi RTT fingerprinting: An analysis of the properties and the performance in non-line-of-sight environments’ is illustrated in Figure 1. It investigates the effectiveness of Wi-Fi-based indoor location systems, focusing specifically on the use of Wi-Fi round-trip time (RTT) fingerprinting in complex indoor environments. Through rigorous analysis, the study shows that Wi-Fi RTT fingerprinting has superior performance compared to conventional methods. It achieves an accuracy of 0.6 metres, proving its potential as a robust and accurate indoor positioning solution. These results contribute to the growing scientific knowledge in the field of indoor positioning and provide valuable insights for further advances in this area.

Figure 1. “WiFi RTT fingerprinting: an analysis of the properties and the performance in non-line-of-sight environments”, generated with the help of Bing image creator, Juli 2023.

In the contribution, ‘Exploring the potential of social media to study environmental topics and natural disasters’, Kenzo Milleville, Samuel Van Ackere, Jana Verdoodt, Steven Verstockt, Philippe De Maeyer & Nico Van de Weghe, document the growing importance of social media as a key communication medium (illustration in Figure 2). They demonstrate the immense potential for gaining valuable insights through large-scale data processing. The main objective is to develop and implement a robust pipeline for the automatic extraction and analysis of Twitter (now named X) data on natural disasters and environmental issues. This approach aims to provide an additional layer of spatio-temporal data that can be used to study the immediate and lasting impacts of natural disasters, climate change, and environmental phenomena on a global scale. Their preliminary analysis, which focused on wildfires, was conducted in four different languages, confirming the need for multilingual support to enable a comprehensive global analysis. Interestingly, the results showed a positive correlation between wildfire occurrence and tweeting behaviour, as well as the geographic spread of fires. In addition, simple sentiment predictions were found to be of limited value when data were aggregated on a large scale (Figure 3). A refined sentiment detection model was then used to detect sentiment in tweets about nuclear energy. Encouragingly, this method yielded promising results. Future research efforts aim to expand the dataset and develop customised models to enable more in-depth analysis of the global impact of natural disasters and environmental issues.

Figure 3. “Influence of tracking duration on the privacy of individual mobility graphs”, generated with the help of Bing image creator, Juli 2023.

Figure 2. “Exploring the potential of social media to study environmental topics and natural disasters”, generated with the help of Bing image creator, Juli 2023.

The paper ‘Influence of tracking duration on the privacy of individual mobility graphs’ by Nina Wiedemann, Henry Martin, Esra Suel, Ye Hong & Yanan Xin examines the re-identification risks associated with location graphs, compact representations of human mobility that protect privacy. Despite the privacy benefits, the study shows that these graphs can still be used to re-identify users based on topology. The research focuses on how the duration of tracking affects the risk of re-identification. Using a one-year dataset of 137 users divided into subsets with different durations, the experiment yields a re-identification accuracy of 0.41% to 20.97%, depending on the duration of tracking the pool and the test users. The study shows that re-identification is affected by the duration of the pool and the tracking of the test users, with greater similarity in duration increasing the risk. Sociodemographic characteristics have minimal impact, while mobility and graphical characteristics play a role. The results highlight the importance of tracking duration for user privacy and recommend limiting duration or resetting user IDs periodically when long-term tracking data is stored.

Figure 4 illustrates the paper ‘A Mobile Mapping Solution for VRU Infrastructure Monitoring via Low-Cost LiDAR Sensors’. As documented by authors Johanna Vogt, Mario Ilic & Klaus Bogenberger, LiDAR scanning is a widely used method for assessing the condition of roads and vehicle infrastructure. However, it overlooks pedestrian and bicycle infrastructure and can be expensive. Recent advances in electronics have made it possible to incorporate powerful LiDAR sensors into personal devices such as smartphones and tablets. This study introduces a new category of wearable LiDAR systems specifically designed to map infrastructure for vulnerable road users at a smaller scale. In the study, qualitative experiments were conducted using a wearable LiDAR system and a 3D scanner app to achieve low-cost mobile mapping of infrastructure for vulnerable road users. The results showed that the system produced detailed and qualitative point clouds at low speeds that allowed accurate identification of potholes and bumps. The capabilities of this sensor open up various mobile mapping solutions for vulnerable road user infrastructure monitoring, including crowd-sourcing options for urban data mining and interfaces for urban digital twins, all at a significantly lower cost than traditional sensor solutions.

Figure 4. “A mobile mapping solution for VRU infrastructure monitoring via low-cost LiDAR-Sensors”, generated with the help of Bing image creator, Juli 2023.

The study ‘Augmented Reality Landmarks on Windshield and Their Effects on the Acquisition of Spatial Knowledge in Autonomous Vehicles’ by Rui Li investigates the effects of augmented reality landmarks on spatial learning in autonomous vehicles. A generated illustration is shown in the picture in Figure 5. The study aims to improve drivers’ spatial knowledge by extending the visualisation of distant landmarks to the windshield using augmented reality technology. An experiment was conducted using simulated videos of autonomous driving to evaluate the effects of augmented reality landmarks and road conditions on spatial knowledge acquisition. The results show that distant augmented reality landmarks improve the efficiency of route and direction knowledge acquisition. However, the effect of augmented reality landmarks on route knowledge on local roads is not significant. Configuration knowledge is influenced more by the type of road than by augmented reality landmarks (Figure 6). The study highlights the need for further research to symbolise additional information in augmented reality landmarks to support spatial learning and evaluate their effects on configuration knowledge.

Figure 6. “Analytics of historical human migration patterns: use cases of Amsterdam and Copenhagen”, generated with the help of Bing image creator, Juli 2023.

Figure 5. “Augmented reality landmarks on windshield and their effects on the acquisition of spatial knowledge in autonomous vehicles”, generated with the help of Bing image creator, Juli 2023.

In their paper, ‘Analytics of historical human migration patterns: use cases of Amsterdam and Copenhagen’, Irma Kveladze, Marina Georgati, Carsten Kessler & Henning Sten Hansen examine the relationship between human migration and the emergence of extensive data streams that serve as valuable sources of information. The authors present a novel method of analysis aimed at understanding patterns of human migration in urban contexts, focusing specifically on Amsterdam and Copenhagen. Unlike alternative approaches, their proposed method consists of mapping migration data within grid cells while retaining key data attributes. Through the use of statistical computations, the authors improve the visualisation and analysis capabilities of the data, allowing for a more detailed examination of the dynamics of human migration in these urban areas (Figure 7).

Figure 7. Prompt - “towards human and context-adaptive mobile geographic information displays to support spatial learning for pedestrian navigation”, generated with the help of Bing image creator, September 2023.

In her paper ‘Neuroadaptive LBS: Towards human–, context–, and task-adaptive mobile geographic information displays to support spatial learning for pedestrian navigation’, Sara Fabrikant, argues that structured neuro-adaptive mobile geographic information displays have the potential to significantly improve the daily lives of a wide range of people. They regularly face the need to make time-sensitive and socially relevant decisions while on the move. Questions remain: What cognitive processes underlie human decision making when guided by human- and context-dependent neuro-adaptive mobile geographic information displays? In this ongoing empirical investigation, the authors seek to address these fundamental questions. They also present innovative methodological approaches aimed at assessing the effects of perceptual, neurocognitive, psychophysiological, and display design factors on decision making and spatial learning in pedestrian mobility, considering a wide range of users and mobility contexts.

The numerous researches and publications have shown that Location Based Services has become ubiquitous in daily life. This selection of papers is an example of the research efforts to realise the 4A vision of Location Based Service: ‘anywhere, anytime, for anyone and anything’. We hope you find these contributions informative, interesting, and stimulating.

In this editorial, we have created a visual impression for each post using AI tools to help readers understand the upcoming open research questions related to Artificial Intelligence enabled Location Based Services: What does Artificial Intelligence mean for Location Based Services? Can we develop a generic Artificial Intelligence model, that can be used for different Location Based Services application scenarios after pre-training? How can we consider essential ethical aspects like fairness and transparency? The 2022 conference on Location Based Services, followed by this special issue, would not have been possible without the professional help of our scientific advisory board, the staff and colleagues of the University of Augsburg, Geoinformatics, and the Technical University Munich (TUM), Cartography and Visual Analytics. Many thanks for this. The next LBS conference (LBS2023) is already being planned and will be hosted by the University of Ghent.