Poster (2017, November 27)
Human mobility has opened up to many themes in recent years. Human behavior and how a driver might react to certain situations, whether dangerous (e.g. an accident) or simply part of the evolution of new technologies (e.g. autonomous driving), leaves many avenues to be explored. Although experiments have been deployed in real situations, it remains difficult to encounter the conditions that certain studies may require. For this reason, we have set up a driving simulator (comprising several modules) that is able to reproduce a realistic driving environment. Although, as the literature has already demonstrated, the conditions are often far from reality, simulation platforms are nonetheless capable of reproducing an incredibly large number of scenarios on the fly. In this poster, we explain how we conceived the simulator, as well as the system we developed for collecting metrics on both the driver and the simulation environment. In addition, we take advantage of this conference to publicly share a dataset consisting of 25 drivers performing the same road circuit on the "Project Cars" game.
In recent years, multimodal transportation has become a challenging approach to route planning. Most existing planning systems usually rely on data sourced from different organisations, enabling the user to select a limited number of routing strategies. As part of the MAMBA project, developed in Luxembourg until 2017, we have been interested in the potential benefits of multimodal mobility systems. A key factor has been integrated into our studies: the need for a personalised experience at user level, whether when selecting the means of transport or describing user habits (e.g. route style, environment). In this context, we have developed a platform for planning personalised multimodal trips, broken down into the three main modules presented in this demonstration. More importantly, this platform has been developed to facilitate the daily mobility of people in Luxembourg, and considers datasets and characteristics that are specific to this region, which has an exceptionally high volume of daily commuting between Luxembourg and neighbouring countries.
Scientific Conference (2017, October)
Within the world of wireless technologies, Bluetooth has recently been at the forefront of innovation. It is becoming increasingly relevant for vehicles to become aware of their surroundings. Therefore, having knowledge of nearby Bluetooth devices, both inside and outside other vehicles, can provide the listening vehicles with enough data to learn about their environment. In this paper, we collect and analyze a dataset of Bluetooth Classic (BC) and Low Energy (BLE) discoveries. We evaluate their respective characteristics and ability to provide context-aware information from a vehicular perspective. By taking a look at data about the encountered devices, such as GPS location, quantity, quality of signal and device class information, we infer distinctive behaviors between BC and BLE relative to context and application. For this purpose, we propose a set a features to train a classifier for the recognition of different driving environments (i.e. road classes) from Bluetooth discovery data alone. Comparing the performance of our classifier with different sampling parameters, the presented results indicate that, with our feature selection, we are able to predict with reasonable confidence up to three classes (Highway, City, Extra-Urban) by using only discovery data and no geographical information. This outcome gives promising results targeted at low energy and privacy-friendly applications and can open up a wide range of research directions.
in ISPRS International Journal of Geo-Information (2017), 6(3), 62
The potential of geospatial big data has been drawing attention for a few years. Despite the larger and larger market penetration of portable technologies (nomadic and wearable devices like smartphones and smartwatches), their opportunities for travel behavior analysis are still relatively unexplored. The main objective of our study is to extract the human mobility patterns from GPS traces in order to derive an indicator for enhancing Collaborative Mobility (CM) between individuals. The first step, extracting activity duration and location, is done using state-of-the-art automated recognition tools. Sensors data are used to reconstruct individual’s activity location and duration across time. For constructing the indicator, in a second step, we defined different variables and methods for specific case studies. Smartphone sensor data are being collected from a limited number of individuals and for one week. These data are used to evaluate the proposed indicator. Based on the value of the indicator, we analyzed the potential for identifying CM among groups of users, such as sharing traveling resources (e.g., carpooling, ridesharing, parking sharing) and time (rescheduling and reordering activities).
The study of human mobility and activities has opened up to an incredible number of studies in the past, most of which included the use of sensors distributed on the body of the subject. More recently, the use of smart devices has been particularly relevant because they are already everywhere and they come with accurate miniaturized sensors. Whether it is smartphones, smartwatches or smartglasses, each device can be used to describe complementary information such as emotions, precise movements, or environmental conditions. In this short paper, we release the applications we have developed and an example of a collected dataset. We propose that opening multi-sensors data from daily activities may enable new approaches to studying human behavior.
in International Journal of Distributed Sensor Networks (2017), 13(8),
Recent technological advances and the ever-greater developments in sensing and computing continue to provide new ways of understanding our daily mobility. Smart devices such as smartphones or smartwatches can, for instance, provide an enhanced user experience based on different sets of built-in sensors that follow every user action and identify its environment. Monitoring solutions such as these, which are becoming more and more common, allows us to assess human behavior and movement at different levels. In this article, extended from previous work, we focus on the concept of human mobility and explore how we can exploit a dataset collected opportunistically from multiple participants. In particular, we study how the different sensor groups present in most commercial smart devices can be used to deliver mobility information and patterns. In addition to traditional motion sensors that are obviously important in this field, we are also exploring data from physiological and environmental sensors, including new ways of displaying, understanding, and analyzing data. Furthermore, we detail the need to use methods that respect the privacy of users and investigate the possibilities offered by network traces, including Wi-Fi and Bluetooth communication technologies. We finally offer a mobility assistant that can represent different user characteristics anonymously, based on a combination of Wi-Fi, activity data, and graph theory.
Poster (2016, December 08)
Recent technological advances in communication technology have provided new ways to understand human mobility. Connected vehicles with their rising market penetration are particularly representative of this trend. They become increasingly interesting, not only as sensors, but also as participants in Intelligent Transportation System (ITS) applications. More specifically, their pervasive connectivity to cellular networks enables them as passive and active sensing units. In this paper, we introduce LuST-LTE, a package of open-source simulation tools that allows the simulation of vehicular traffic along with pervasive LTE connectivity.
in Proceedings of the 14th ACM Conference on Embedded Networked Sensor Systems (SenSys 2016) (2016, November 14)
Human mobility is one of the key topics to be considered in the networks of the future, both by industrial and research communities that are already focused on multidisciplinary applications and user-centric systems. If the rapid proliferation of networks and high-tech miniature sensors makes this reality possible, the ever-growing complexity of the metrics and parameters governing such systems raises serious issues in terms of privacy, security and computing capability. In this demonstration, we show a new system, able to estimate a user's mobility profile based on anonymized and lightweight smartphone data. In particular, this system is composed of (1) a web analytics platform, able to analyze multimodal sensing traces and improve our understanding of complex mobility patterns, and (2) a smartphone application, able to show a user's profile generated locally in the form of a spider graph. In particular, this application uses anonymized and privacy-friendly data and methods, obtained thanks to the combination of Wi-Fi traces, activity detection and graph theory, made available independent of any personal information.
in Proceedings of the 2016 International Conference on Information and Communications Technology Convergence - ICTC 2016 (2016, October)
Recent technological advances and the ever-greater developments in sensing and computing continue to provide new ways of understanding our daily mobility. Smart devices such as smartphones or smartwatches can, for instance, provide an enhanced user experience based on different sets of built-in sensors that follow every user action and identify its environment. Monitoring solutions such as these, which are becoming more and more common, allow us to assess human behavior and movement at different levels. In this article, we focus on the concept of human mobility. With the participation of 13 individuals, we carried out an experiment to discover how groups of sensors currently available in smartphones and smartwatches can help to distinguish different profiles and patterns of human mobility. We show that it is possible to use not only motion sensors but also physiological sensors and environmental data provided, for instance, by Wi-Fi. Finally, detailed study of these categories enables us to offer a way of representing the mobility of individual users, based on anonymized traces and graph theory.
in Journal of Wireless Mobile Networks, Ubiquitous Computing, and Dependable Applications (2016), 7(3),
The rapid emergence of new technologies in recent decades has opened up a world of opportunities for a better understanding of human mobility and behavior. It is now possible to recognize human movements, physical activity and the environments in which they take place. And this can be done with high precision, thanks to miniature sensors integrated into our everyday devices. In this paper, we explore different methodologies for recognizing and characterizing physical activities performed by people wearing new smart devices. Whether it’s smartglasses, smartwatches or smartphones, we show that each of these specialized wearables has a role to play in interpreting and monitoring moments in a user's life. In particular, we propose an approach that splits the concept of physical activity into two sub-categories that we call micro- and macro-activities. Micro- and macro-activities are supposed to have functional relationship with each other and should therefore help to better understand activities on a larger scale. Then, for each of these levels, we show different methods of collecting, interpreting and evaluating data from different sensor sources. Based on a sensing system we have developed using smart devices, we build two data sets before analyzing how to recognize such activities. Finally, we show different interactions and combinations between these scales and demonstrate that they have the potential to lead to new classes of applications, involving authentication or user profiling.
in Proceedings of the 2016 IEEE International Smart Cities Conference (ISC2) (2016, September)
Today’s mobile penetration rates enable cellular signaling data to be useful in diverse fields such as transportation planning, the social sciences and epidemiology. Of particular interest for these applications are mobile subscriber dwell times. They express how long users stay in the service range of a base station. In this paper, we want to evaluate whether dwell time distributions can serve as predictors for road travel times. To this end, we transform floating car data into synthetic dwell times that we use as weights in a graph-based model. The model predictions are evaluated using the floating car ground truth data. Additionally, we show a potential link between handover density and travel times. We conclude that dwell times are a promising predictor for travel times, and can serve as a valuable input for intelligent transportation systems.
in IEEE Transactions on Vehicular Technology (2016), 65(7), 5720-5725
In a near future, wireless networks will be one of the key technologies for road traffic management in smart cities. Vehicles and dedicated roadside units should be interconnected through wireless technologies such as IEEE 802.11p (WAVE). Traffic light and road signs may also take their place in this architecture, forming a large-scale network of small devices that report measurements, take orders from a control center, and are able to take decisions autonomously based on their local perception. Such a network shares many similarities with classical wireless sensor and actuator networks, starting with its distributed organization and with the role of the control center. However, its topology, and subsequently the appropriate selection of protocols and algorithms, will be strongly influenced by each city's characteristics. In this article, we characterize and discuss probable topologies of these networks. The aim of this work is to provide network models that can be used to evaluate protocols and algorithms using realistic scenarios in place of generic random graphs. We deploy such networks over 52 city maps extracted from OpenStreetMap and characterize the resulting graphs, with a particular focus on connectivity aspects (degree distribution and connected components). The tools, the complete datasets, OMNeT++ network models are available freely online.
in Proceedings of the 9th International Conference on Human System Interaction (HSI'16) (2016, July)
Smart and wearable devices are trendy electronic objects that have become increasingly popular in recent years. Those devices are, by definition, tightly connected with the user's personal activities. Authentication is therefore a critical feature for both identifying users and personalizing the services on the device. In particular, the emergence of smartglasses changed the way we thought a wearable could assist users in their daily activities. As designed by commercial providers, smartglasses are sold with a very specific set of interactions capabilities. These capabilities have a strong impact on how comfortably or safely users may authenticate to their smartglasses. For this reason, we investigate in this paper the different authentication methods available for smartphones and we comparatively position the smartglasses in the design space of authentication methods. We propose a new approach based on touch input on an arbitrary surface using thermal camera input. This approach aims to circumvent the lack of touch surface provided by smartglasses, while maintaining an acceptable level of security.
in Proceedings of the 9th International Conference on Human System Interaction (HSI'16) (2016, July)
Despite the rapid pace of gadgets released on the market, research in the area of usable interfaces for wearables is lagging behind. Smartglasses are new wearables that embed diverse sensors but also have small displays, and this makes it hard for the wearer to visualize real-time data. To bridge this gap, the contribution of this paper is threefold. First, we propose a data representation model to combine applications and services that match user activities and contexts. Second, we present an approach of showing relevant services to the user based on 'tiles' (such as those in recent Microsoft Windows interfaces) while considering the device constraints. Finally, we suggest that combining those two aspects can open the way to personalized services for the end user, creating new ways of interacting with applications and devices.
Presentation (2016, June 30)
in Proceedings of the 7th IEEE INFOCOM International Workshop on Mobility Management in the Networks of the Future World (2016, April)
The emergence of new connected devices has opened up new opportunities and allowed to imagine concepts that bring computer sciences and social sciences closer together. In particular, today's increasingly sophisticated miniature sensors allow to track and understand human activities and behavior with a great precision. Taking different approaches and perspectives, we use in this paper smartwatches and smartglasses to explore these behaviors and show that these objects, considered by many as gadgets, have an important role to play in understanding the lives of individuals. The main objective of this work is to introduce two new scales of activity detection, which lacks a formal and consistent definition in the literature. First, we propose a model that precisely detects and interprets movements made by a person wearing smart devices. Then, we use this model to show different interactions between those micro-activities and bigger chunks of behaviors we call macro-activities. Using a new concept based on 3D visualization, we finally show that combining those two scales and using a limited dataset, it is possible to distinguish between different individuals when they are performing very similar activities. The findings of this study lead the way to enhanced user profiling.
in IEEE Intelligent Transportation Systems Magazine (2016)
Both the industrial and the scientific communities are working on problems related to vehicular traffic congestion, intelligent transportation systems, and mobility patterns using information collected from a variety of sources. Usually, a vehicular traffic simulator, with an appropriate scenario for the problem at hand, is used to reproduce realistic mobility patterns. Many mobility simulators are available, and the choice is made based on the type of simulation required, but a common problem is finding a realistic traffic scenario. The aim of this work is to provide and evaluate a scenario able to meet all the basic requirements in terms of size, realism, and duration, in order to have a common basis for evaluations. In the interest of building a realistic scenario, we used information from a real city with a typical topology common in mid-size European cities, and realistic traffic demand and mobility patterns. In this paper, we show the process used to build the Luxembourg SUMO Traffic (LuST) Scenario, and present a summary of its characteristics together with our evaluation and validation of the traffic demand and mobility patterns.
in Proceedings of the 22nd IEEE Symposium on Communications and Vehicular Technology in the Benelux (SCVT 2015) (2015, November 24)
Recent advances in the field of intelligent transportation systems have focused on the use of wireless networks to link vehicles and road infrastructure. Applications that might result from such networks range from the adaptive management of traffic lights to the detection of traffic jams and accidents. Whatever the case may be, it seems important to explore the possibilities and limitations of such networks, which the literature often portrays in a somewhat idealistic way (e.g. no packet loss, fully connected sensors, etc.). In this paper, we study the deployment of wireless sensor networks at intersections in some of the world's major cities and characterize their topologies. Using a propagation model that corresponds to a 2.4GHz IEEE 802.15.4 network interface, we focus our study on the global connectivity of graphs resulting from different networks. By deploying this type of network over 52 city and region maps extracted from OpenStreetMap, we show that cities can reasonably be classified into three network structure categories of low connectivity (i.e. a high number of connected components) and that it should be feasible to improve the networks by adding sensors. All the tools and the complete dataset are freely available online.
in Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering - Mobile Computing, Applications, and Services (2015, November)
The continuous development of new technologies has led to the creation of a wide range of personal devices embedded with an ever increasing number of miniature sensors. With accelerometers and technologies such as Bluetooth and Wi-Fi, today's smartphones have the potential to monitor and record a complete history of their owners' movements as well as the context in which they occur. In this article, we focus on four complementary aspects related to the understanding of human behaviour. First, the use of smartwatches in combination with smartphones in order to detect different activities and associated physiological patterns. Next, the use of a scalable and energy-efficient data structure that can represent the detected signal shapes. Then, the use of a supervised classifier (i.e. Support Vector Machine) in parallel with a quantitative survey involving a dozen participants to achieve a deeper understanding of the influence of each collected metric and its use in detecting user activities and contexts. Finally, the use of novel representations to visualize the activities and social interactions of all the users, allowing the creation of quick and easy-to-understand comparisons. The tools used in this article are freely available online under a MIT licence.
in Proceedings of the 7th International Conference on Mobile Computing, Applications and Services (MobiCASE '15) (2015, November)
SWIPE is a platform for sensing, recording and processing human dynamics using smart devices. The idea behind this type of system, which exists for the most part on smartphones, is to consider new metrics from wearables - in our case smartwatches. These new devices, used in parallel with traditional smartphones, provide clear indicators of the activities and movements performed by the users who wear them. They can also sense environmental data and interactions. The SWIPE architecture is structured around two main elements, namely (1) an Android application deployed directly on the devices, allowing them to synchronize and collect data; and (2) a server for storing and processing the data. This publication is intended to communicate on the platform with both the scientific and the industry communities. SWIPE is freely distributed under a MIT license.
in Proceedings of the 13th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys '15) (2015, May)
Recent technological advances have allowed the development of miniaturized sensors and the emergence of a wide range of connected objects. Whether it's smartphones or in the broader sense wearables, the diversity of these devices and their accessibility opens up new fields for applications in the computer sciences. Smartwatches, which are experiencing a boom on the market, will be integral to the research that will shape the Internet in the years to come, namely big data, sensing systems and human behavior. Our demonstration falls within this context and aims to demonstrate the potential of these emerging technologies to respond to problems and to way of thinking introduced by industry and the scientific community, which are generally limited to smartphone sensing frameworks. Further, we plan to present our research platform, SWIPE, which is dedicated to collecting, studying and learning about human dynamics by means of an ecosystem of wearables.