According to the world report in 2019 about vision health presented by the World Health Organization (WHO), there are 2.2 billion people in the world with some kind of visual impairment. Furthermore, in a recent National Health Interview Survey report, 25.5 million adult Americans 18 and older reported experiencing vision loss. Of these 25.5 million American adults, 15.3 million women and 10.1 million men report experiencing significant vision loss. As a result, this population is constantly at risk of a fall and its consequences. This paper presents the first module of our fall prevention system for visually impaired people. The module corresponds to a system of floor and path recognition in outdoor environments. At the core of the system is a Mask R-CNN trained with around 40.000 outdoor images. The proposed approach has reached a performance in a combination of training and testing of 92%, using a combined set of indoor and outdoor environment images.

Vehicular Crowd-sensing (VCS) is a new data collection paradigm that leverages the unique characteristics of vehicular mobility to collect sensing data. One of the main challenges for VCS is how to assign sensing tasks among participant so as to maintain the required Quality of Sensing Data (QoSD), while keeping participation profitable. We tackle this challenge by designing an incentive mechanism that encourages the collection of high QoSD, while improve participant utilities. The proposed mechanism includes a platform which post a set of sensing tasks and the associated rewards, and a set of participant vehicles equipped with sensors, who compete for these rewards. We model this competition as a non-cooperative game in which the set of vehicles are the players, and their trajectories are their strategies. Using open-street maps, SUMO vehicular traffic simulator, and extensive simulations, we show that our algorithm significantly outperforms a greedy approach in terms of QoSD, average vehicle utility, spatial coverage, and road utilization.

Monitoring environmental variables in lower layers of the atmosphere is an important activity to measure changes that result from natural events, and human interventions. Volcano eruptions, commercial aviation, and the massive spread of pesticides using light aircraft are just some examples of low layer atmosphere polluters. Twice a day, every day of the year, weather balloons are released simultaneously from almost 900 locations worldwide to monitor environmental variables. The flight of these synthetic rubber balloons last for around 2 hours, then they become pollution too. Recent advances in small unmanned aerial vehicles (UAVs) with built in sensors, and their emerging role in business supply chain make UAVs ideal participants for environmental monitoring. In this paper, we present an incentive mechanism for UAV-Crowdsensing. The core of the proposed mechanism consists of a recurrent reverse action and a recruitment model. By these two components, the system encourages UAVs sensing from locations that maximize volume coverage within a given budget. Through extensive simulations, we evaluate the performance of the proposed incentive mechanism.

A recent study showed that around 77% of the U.S. population experiences regularly physiological symptoms of stress such as sleep deprivation, fatigue, and migraines, and 73% psychological symptoms such anxiety, angry, and lack of focus. Thus, psychological stress remains an important issue of public health. In this paper, propose a mechanism for the selection of stress interventions based on fuzzy control theory. Here, we model the user as a dynamical system, in particular, we use the well-known analogy of a fluid reservoir or tank to model the user's accumulation of stress. In this model, inputs correspond to stressors which increase the level of fluid and the outputs correspond to interventions which release stress and drain the tank. Thus, we tackle the problem keeping a user's healthy stress level by using a fuzzy logic approach. Through several examples and extensive simulations we evaluate the performance of our proposed mechanism for the selection of stress interventions.

In this paper, we design an incentive mechanism for data collection in smart cities. We propose an incentive mechanism for crowdsensing with multiple crowdsourcers. We model the incentive mechanism as a noncooperative game. We consider two different pricing mechanisms when the crowdsourcers fixed the rewards in advance, and when the crowdsourcers dynamically set the rewards in order to maximize their own utilities. A discrete time dynamic inspired by the well known best response dynamic, called elite strategy dynamics, is proposed to compute a Nash equilibrium of the modeled game. Comprehensive simulations were presented to evaluate the performance of the proposed incentive mechanism.

Recent studies have modeled the incentive mechanism as a complete information game where the contributors have common knowledge. However, that assumption is not realistic in real world scenarios. In this paper, we present an incentive mechanism for CS in sealed markets in which participants have incomplete information on other participants' behavior. An iterative game framework is introduced where the solution is achieved after a number of iterations. We also address the question “do uncoordinated contributors converge to an equilibrium?. In fact, we are concerned with the convergence of the contributors to an equilibrium under a natural dynamics. The well-known best response dynamics with different rules of play is studied. In addition, a strategy for the platform to assign the optimal budget for the initial state of the market is presented. Through theoretical analyses and extensive simulations, we also evaluate the performance of the proposed incentive mechanism.

Crowdsensing (CS) is a new data collection paradigm based on the willingness of people to utilize their mobile devices to sense and transmit data of interest. Given the large amount of cellular users, mobile sensor networks will be able to collect enough data to address large-scale societal problems in a fast, easy, and cost-effective manner. One important issue in CS is that of privacy; without appropriate privacy-preserving mechanisms, many users will not be willing to participate in the data collection process. This paper presents the state-of-the-art in privacy-preserving mechanisms for CS systems. After a general description of CS systems and their main components, this paper addresses the most important issues to consider in the design, implementation, and evaluation of privacy-preserving mechanisms. Then, following a new taxonomy, the most important mechanisms available in the literature are described and qualitatively evaluated. Finally, this paper presents research challenges that should be addressed in order to improve the performance of future privacy-preserving mechanisms for CS systems.

Mobile context recognition attempts to infer the context of a mobile phone user. Machine learning algorithms exhibit high classification accuracy in these applications. Most approaches are very power-inefficient because they record data from all sensors at all time. Intelligently cycling sensors could greatly improve the power efficiency of context recognition services. We propose a decision tree-based machine learning algorithm which optimizes not only on classification accuracy, but also on data retrieval costs based on power efficiency. We show that in a simple physical activity recognition application, the use of this new algorithm results in a significant decrease in power consumption while maintaining a high classification accuracy.

This paper examines just-in-time adaptive interventions (JITAIs) for stress, a pervasive and affective computing application with significant implications for long-term health and quality of life. We discuss fundamental components needed to enabling JITAIs based for one kind of affect data stress. Chronic stress has significant long-term behavioral and physical health consequences, including an increased risk of cardiovascular disease, cancer, anxiety and depression. This paper conducts post-hoc experiments and simulations to demonstrate feasibility of both real-time stress forecasting and stress intervention adaptation and optimization. Using physiological data collected by ten individuals in the natural environment for one week, we show 1) that simple Hidden Markov Models (HMMs) can be used to forecast physiological measures of stress with up to 3 minutes in advance; and 2) Q-Learning (QL) combined with eligibility traces could be used by an affective computing system to adapt and deliver any number and type of interventions in response to changes in affect. Our hope is that this work will take us one step closer to using pervasive devices to assist in the daily management of chronic stress and other affect-related challenges.

Crowd sensing (CS) is an approach to collecting many samples of a phenomena of interest by distributing the sampling across a large number of individuals. While any one individual may not provide sufficient samples, aggregating samples across many individuals provides high-quality, high-coverage measurements of the phenomena. Thus, for participatory sensing to be successful, one must motivate a large number of individuals to participate. In this work, we review a variety of incentive mechanisms that motivate people to contribute to a CS effort. We then establish a set of design constraints or minimum requirements that any incentive mechanism for CS must have. These design constrains are then used as metrics to evaluate those approaches and determine their advantages and disadvantages. We also contribute a taxonomy of CS incentive mechanisms and show how current systems fit within this taxonomy. We conclude with the identification of new types of incentive mechanisms that require further investigation.

Participatory sensing (PS) systems rely on the willingness of mobile users to participate in the collection and reporting of data using a variety of sensors either embedded or integrated in their cellular phones. However, this new data collection paradigm has not been very successful yet mainly because of the lack of incentives for participation. Although several incentive schemes have been proposed to encourage user participation, none has used location information and imposed budget and coverage constraints, which will make the scheme more realistic and efficient. We propose a recurrent reverse auction incentive mechanism with a greedy algorithm that selects a representative subset of the users according to their location given a fixed budget. Compared to existing mechanisms, our incentive scheme improves the area covered by more than 60 percent acquiring a more representative set of samples after every round while maintaining the same number of active users in the system and spending the same budget.