Rita Tse

Challenges and opportunities in immersive vehicular sensing: Lessons from urban deployments

Vehicles provide an ideal platforms for a plethora of emerging applications such as networked gaming, multimedia content delivery and urban sensing. Cars have no power constraints and they can be instrumented with high end computational units and graphic devices. The deployment at scale of urban vehicular systems, however, requires a careful design able to consider challenges across several domains. Vehicular systems are arguably a prominent example of cyber-physical systems. The development of such systems requires a truly multidisciplinary approach and a close integration between the application, communication, and physical domains. Hardware and software mounted on vehicles will face a harsh physical and communication environment that will greatly affect all the system components. In this paper we report on the challenges and opportunities for multimedia vehicular urban sensing systems based on our field experiences in Macao (China) and Los Angels (USA). We designed and built the components for a pollution monitoring system able to support closed-loop optimization between pollution and traffic management. Our initial set of prototype vehicles are now running in the city of Macao and they are measuring the air parameters as well as the urban traffic. The paper aims at exposing some of the issues encountered, outlining the problems of a city wide deployment, and augmenting our in-field experience with the results from large scale simulation studies.

Using geosocial search for urban air pollution monitoring

Abstract While Twitter and other Online Social Networks (OSNs) or microblogs are considered as a source of information for breaking news or uproarious and unexpected events, they could also be exploited as a dense worldwide sensors network for physical measurements. The corpus of geotagged posts from OSNs includes people’s feedbacks about a wide range of topics, with precise temporal and geographical metadata, that can be used as a support or an improvement to hardware sensors. For instance, if collocated people, independently and at the same time, write posts complaining about high temperatures, it could effectively denote a raise of heat in that place. In this paper, we explore the feasibility to use a geographical search on social networks, that is, a geosocial search, about air pollution related posts, as effective air impureness measurements. We evaluate our assumption in large cities over three continents of the planet, where a minimum increment about the number of air pollution related posts in an area, indeed corresponds to a raise of minimum pollution values in such area. Such a correlation can be exploited to integrate and extend existing air pollution monitoring networks. At the end of the manuscript we propose to further employ the time series of posts returned by the geosocial search to predict next pollution values.

Social Network Based Crowd Sensing for Intelligent Transportation and Climate Applications

In recent years, the growing prevalence of social networks makes it possible to utilize human users as sensors to inspect city environment and human activities. Consequently, valuable insights can be gained by applying data mining techniques to the data generated through social networks. In this work, a practical approach to combine data mining techniques with statistical analysis is proposed to implement crowd sensing in a smart city. A case study to analyze the relationship between weather conditions and traffic congestion in Beijing based on tweets posted on Sina Weibo platform is presented to demonstrate the proposed approach. Following the steps of data pre-processing and topic determination, we applied Granger Causality Test to study the causal relationships between weather conditions, traffic congestion and human outdoor activity. The mediation analysis is also implemented to verify human outdoor activity as a mediator variable significantly carrying the influence of good weather to traffic congestion. The result demonstrates that outdoor activity serves as a mediator transmitting the effect of good weather on traffic congestion. In addition, the causes of negative emotion are also studied.

Sensing Pollution on Online Social Networks: A Transportation Perspective

Transportation policy and planning strategies, as well as Intelligent Transportation Systems (ITS), can all play important roles in decreasing pollution levels and their negative effects. Interestingly, limited effort has been devoted to exploring the potential of social network analysis in such context. Social networks provide direct feedback from people and, hence, potentially valuable information. A post telling how a person feels about pollution at a given time at a given location, could be useful to policy-makers, planners or environmentally-aware ITS designers. This work verifies the feasibility of sensing air pollution from social networks and of integrating such information with real sensors feeds, unveiling how people advertise such phenomenon, acting themselves as smart objects, and how online posts relate to true pollution levels. This work explores a new dimension in pollution sensing for the benefit of environmental and transportation research in future smart cities, confronting over 1,500,000 posts and pollution readings obtained from governmental on-the-field sensors over a one-year span.

Enabling street-level pollution and exposure measures: a human-centric approach

In the last several years the awareness about pollution levels and other natural hazards has risen significantly. Millions of citizens globally pay daily attention to air quality and other factors such as for instance UV index and pollens. This new need has been so far met by a plethora of smartphone applications that leverage data collected by the city-operated stations or, in some cases, community-based stations located at the user premises. In this work we exploit recent advances in micro-controllers and sensing technologies to design and implement a personal pollution awareness system tiny enough to be embedded into user accessories. In particular, we will introduce the concept and the architecture of our prototype and show the preliminary results we collected using this prototype.

Can we monitor the natural environment analyzing online social network posts? A literature review

Abstract A number of works have addressed the question of assessing the status and the 2 4 quality of the environment through the lens of Online Social Networks (OSNs). These contributions fall in the area of human-centric sensing, area specialized in using what people spontaneously say on social media to detect the occurrence of given events. Research in this area has exhibited interesting results, regardless of the accuracy of sensing operations. In fact, in some cases it is possible to corroborate the information extracted from OSN posts with the ground truth obtained from specialized hardware sensors. In others, the information extracted from OSNs does not reveal true environmental conditions. Nevertheless, OSNs may help shed light on the sensitivity of human beings to a wide variety of environmental phenomena. We here review the work that has been published to this date. In particular, we provide a survey that may benefit both environmental and computer scientists, as this work aims to show where we stand in the understanding of the complex relationship between human beings and the natural environment, when this is mediated by OSNs.

Deployment of vehicular edge clouds: lessons and challenges

The notion of edge computing has gained much attraction in recent years as an enabling technology for smart city and internet of things applications. In this paper we report the system challenges and solutions encountered when designing and deploying the Macao Polytechnic Institute Smart City sensing system. A small fleet that serves as proof-of-concept for a country wide urban sensing system in Macao, S.A.R. We focus our attention on how a careful system design can ensure smooth operations and mitigate the natural tension between fleet owners and smart city operators. The first are keen to maximize the fleet operations and reduce the downtime, the later are interested in using the fleets to harvest high-quality and fine granularity sensor data. In designing the Macao Polytechnic Institute vehicular cloud we approached the design constraints and proposed system solutions to minimize the impact of the sensing platform on the fleet operations.

Crowd Sensing of Weather Conditions and Traffic Congestion Based on Data Mining in Social Networks

In recent years, the growing prevalence of social networks makes it possible to utilize human users as sensors to inspect city environment and human activities. Consequently, valuable insights can be gained by applying data mining techniques to the data generated through social networks. In this work, a practical approach to combine data mining techniques with statistical analysis is proposed to implement crowd sensing in a smart city. A case study to analyze the relationship between weather conditions and traffic congestion in Beijing based on tweets posted on Sina Weibo platform is presented to demonstrate the proposed approach. Following the steps of raw dataset pre-processing, target dataset processing and statistical data analysis, analytic corpus containing tweets related to different weather conditions, traffic congestion and human outdoor activity is selected to test causal relationships by Granger Causality Test. The mediation analysis is also implemented to verify human outdoor activity as a mediator variable significantly carrying the influence of good weather to traffic congestion. The result demonstrates that outdoor activity serves as a mediator transmitting the effect of good weather on traffic congestion.

Mobile Sensing and Beyond in the Information Age: An Experimental Perspective

While fully understood, the opportunities posed to the consumer market by mobile computing platforms have been so far mainly exploited with smartphones and a few other gadgets. This is true for both the exchange of data and personal communication purposes, but also for sensing operations. Efficient sensing operations, however, may not always be performed embedding sensor hardware in smartphones or other commonly used hardware devices (e.g., portable music players, etc.). This emerges when addressing the problem of sensing physical quantities (e.g., air pollution), which can hardly be detected from a sensor mounted on a smartphone closed into a pocket. This paper considers such problem, revisiting current and past experiences with pollution sensing, tracing future directions of work. In fact, a holistic approach is required: the design of sensing systems cannot be separated from the objects embedding them, as successful sensing operation cannot be possible without accounting for those principles capable of making such sensing items not only useful and functional, but also popular and trendy.