Search

In smart cities, urban monitoring systems rely on advanced mobile sensing and learning technologies to track large-scale urban systems and provide efficient urban services in real time. However, the fidelity and amount of sensors deployed at different geo-communities are closely related to their socioeconomic conditions, demographics, and entrenched geographic patterns, causing inequal sensing opportunities across communities. The biased sensing data contain distorted spatio-temporal patterns of undersensed community, inducing unfairness in subsequent algorithmic prediction and decision-making. This work characterize this unfairness propagation chain of sensing - learning - decision-making process. We introduce the first formal mathematical definitions to quantify and decouple community-level unfairness induced by joint cascading effects of sensing inequality and algorithmic bias. Our real-world experiments with vehicular crowdsensing system in Cangzhou, China verifies that sensing inequality, especially community-level gap of sensor fidelity, result in large fairness gap in spatio-temporal data imputation task. Our preliminary results show that sensing inequality amplifies the algorithmic bias. This work is a critical first step in formally defining and understanding unfairness propagation in intelligent spatio-temporal urban monitoring system.

Impostor syndrome is a psychological pattern characterized by doubts in one’s capabilities or achievements despite evident success. In this paper, we demonstrate the potential of a smartphone-aided solution for impostor syndrome detection among high-achieving college students by presenting some initial findings from our 3-week user study. We collected GPS, Bluetooth, light, screen and app usage data along with Clance Impostor Phenomenon Scale (CIPS) survey results from 37 students. We conducted a correlation analysis between CIPS scores and features extracted from sensor data. Then we grouped the features into three categories that reflect the behavioral characteristics of individuals with impostor syndrome, namely social avoidance, sense of obligation, and restlessness. Based on these findings, we suggest further examination of the use of behavioral sensing data in detecting impostor syndrome.

Irregular patterns of Activities of Daily Living (ADLs) are associated with mild cognitive impairment (MCI) of older adults. Measuring the variability of ADL routines using various non-intrusive sensors in smart home environments presents a great opportunity for early diagnosis of MCI. However, existing studies mostly rely on supervised learning approaches to recognize ADLs and measure their variabilities, which requires large efforts in human observation and manual annotation for constructing training datasets for each home environment. In this context, this study proposes an unsupervised hierarchical clustering method to capture ADL clusters and measure their variabilities. In particular, this study focuses on addressing the challenge in employing data from multiple heterogenous sensors. The results show that the proposed method can capture the variability of ADL routines using high-dimensional non-intrusive sensing data.

With the development of low-cost, low-power sensing and communication technologies, there has been growing interest in the IoT for realizing smart cities, in order to maximize the productivity and reliability of urban infrastructure. One of the most representative examples is smart parking system, which consists of parking sensors and backend servers. In existing systems, parking sensors are placed in each parking spot to monitor the status of the parking spot, and then the sensing data is send to backend servers for further processing. Based on these sensor data, the smart parking system can provide some smart services, such as parking spot availability prediction, remote booking and parking guidance. The premise of all these smart services is that the sensor data is accurate and reliable. However, as the sensor ages, the sensor may produce some unreliable data, which brings great challenges to our smart services. In this paper, we present a novel anomaly detection and calibration system in "Suzhou Center", one of the largest most advanced urban complexes in China. Our system uses supervised machine learning algorithms, focusing on capturing spatio-temporal features. The experimental results show that our system can identify most anomalies and improve the accuracy of the sensor.

Due to the spread of COVID-19, we are desired to avoid crowded places including public transportation. Kyushu University has the largest campus in Japan, called "Ito campus", and the population there is about 20,000 in which 23% of students and 46% of staff use a bus for reaching the campus. The lectures in the first half of 2020 have been conducted online, but we plan to resume face-to-face lectures gradually. At that time, we expect the bus stops and buses to be crowded, especially during rush hour. In this paper, we introduce a system, called Itocon, to visualize the human congestion of bus stops around the campus. Itocon aggregates the sensing data from various sensors deployed around the target bus stops, and calculate and visualize the congestion degrees in real-time. Itocon is developed as a web application to avoid requesting the application install. We hope all the people who use a bus change their moving time based on the congestion information for avoiding human crowds. We explain the details and the future prospects of Itocon.

Smart connected home systems bring different privacy challenges to residents. The contribution of this paper is a novel privacy grounded classification of smart connected home systems that is focused on personal data exposure. This classification is built empirically through k-means cluster analysis from the technical specification of 81 commercial Internet of Things (IoT) systems as featured in PrivacyNotIncluded - an online database of consumer IoT systems. The attained classification helps us better understand the privacy implications and what is at stake with different smart connected home systems. Furthermore, we survey the entire spectrum of analyzed systems for their data collection capabilities. Systems were classified into four tiers: app-based accessors, watchers, location harvesters, and listeners, based on the sensing data the systems collect. Our findings indicate that being surveilled inside your home is a realistic threat, particularly, as the majority of the surveyed in-home IoT systems are installed with cameras, microphones, and location trackers. Finally, we identify research directions and suggest some best practices to mitigate the threat of in-house surveillance.

With development of ubiquitous urban sensory infrastructures, large scale human mobility data have various applications and benefits in mobile networking and urban planning. Demands for sharing these sensing data to research communities have proliferated causing concerns over privacy. Traditional privacy solution is to apply comprehensive perturbation at feature level, which may results in significant utility loss. In this paper, we propose a learning-based generative model with designed classifiers to generate realistic synthetic mobility data with privacy protection at label/system property level. With two plug-in classifiers, i.e., privacy classifier and utility classifier, the trade-off between sensitive and non-sensitive mobility properties would be balanced with customization. To validate the performance, we implement our model based on the urban-scale cellphone data from the Shenzhen city as a case study. The resilience of smart attacks is tested and potential applications are pictured for this semi-supervised learning-based model.

One of the major challenges for fully enjoying the power of machine learning is the need for the high-quality labelled data. To tap-in the gold-mine of data generated by IoT devices with unprecedented volume and value, we discover and leverage the hidden connections among the multimodal data collected by various sensing devices. Different modal data can complete and learn from each other, but it is challenging to fuse multimodal data without knowing their perception (and thus the correct labels). In this work, we propose MultiSense, a paradigm for automatically mining potential perception, cross-labelling each modal data, and then improving the learning models over the set of multimodal data. We design innovative solutions for segmenting, aligning, and fusing multimodal data from different sensors. We implement our framework and conduct comprehensive evaluations on a rich set of data. Our results demonstrate that MultiSense significantly improves the data usability and the power of the learning models.

Manufacturing enterprises are facing the challenges of productivity improvement, product and process quality management due to the lack of data interaction and manufacturing process traceability. In modern factory, number of machineries are pipelined into assembly/production lines which may merge and branch to meet the procedure requirements. Each machine on the assembly line generates huge amount of sensing data and manufacturing data. However, many manufacturing system are not ready to manage big data. This paper aims on developing the distributed traceability system with Blockchain technology on IoT devices in order to improve the stability of the factory production line, to reduce defect rate and to bring the operational performance to a new level. Without depending on centralized storage, it is a robust, truly distributed peer-to-peer system and capable of node failure tolerance. With better scalability and data interaction, this distributed ledger network enables the use of data collection with security and potential of public traceability system protocol which allows multiple factories to participate, build an ecosystem of traceability together.

Road traffic accidents are a major challenge in urban transportation systems. An effective countermeasure to address this problem is to accurately forecast the traffic risks in a city before accidents actually happen. Current traffic accident prediction solutions largely rely on accurate data collected from infrastructure-based sensors, which is not always available due to various resource constraints or privacy and legal concerns. In this paper, we address this limitation by exploring social sensing, a new sensing paradigm that uses humans as sensors to report the states of the physical world. In particular, we consider two types of publicly available social sensing data sources: social media data (e.g., traffic posts on Twitter) and open city data (e.g., traffic data from the city web portal). In this paper, we develop the RiskCast, an inductive multi-view learning approach to accurately forecast the traffic risk by exploiting the social sensing data under a principled co-regularization framework. The evaluation results on a real world dataset from New York City show that RiskCast significantly outperforms the state-of-the-art baselines in forecasting the traffic risks in a city.

Wearable sensors have become more commonly used in everyday basis and powerful in terms of computational capacity and sensing resources, including capability to collect data from different bio-signals. The data collected from everyday wearables offers huge opportunities to monitor people's everyday life without expensive laboratory measurements, including also behaviours and conditions only rarely seen in controlled laboratory environments. So far wearable sensors have mostly been used to monitor motion, but bio-sensor powered wearables can do a lot more: they can be used to monitor physiological reactions inside the human body as well as some psycho-physical reactions such as affection and stress. This development enables multiple interesting and important applications, such as early detection of diseases, seizures, and attacks. With stock wearables worn in everyday basis, one of the biggest challenges for such applications is the sensing data itself. In order to train reliable recognition and prediction models, high quality training data with labels needs to be collected. This paper focuses on lessons learned of challenges in data quality and quantity when such data sets are gathered. We discuss our own experiences when collecting data using wearable sensors for early detection of migraine attacks, but the same lessons learned can be generalized to other studies utilizing wearables for recognition medical symptoms and users' everyday behaviour.

Immersive Virtual Reality (IVR) does not afford social cues for communication, such as sweaty palms to indicate stress, as users can only see an avatar of their collaborator. Prior work has shown that this data is necessary for successful collaboration, which is why we propose to augment IVR communication by (1) real-time capturing of physiological senses and (2) leveraging the unlimited virtual space to display these. We present the results of a focus group (N=7) and a preliminary study (N=32) that investigate how this data may be visualized in a playful interaction and the effects they have on the performances of the collaborators.

Mobile crowd sensing (MCS) collects city-scale sensing data with low cost and high efficiency. One important goal of MCS is to ensure high quality of sensing coverage to provide sufficient information to data analysis end. However, the goal of the MCS may be inconsistent with the goal of vehicles. This inconsistency between goals results in a bad sensing coverage and decreases the quality of the collected information. Key challenges to resolve this inconsistency include the heterogeneous target desired spatio-temporal distributions, limited budget constraining the ability to incentivize more taxis, and high computational complexity. This work shows a vehicle dispatching system to optimize the sensing coverage of the sampled data with a limited budget and taxis. Our system models the objective function using the similarity between the collected data distribution and target distribution, and introduces an algorithm combining vehicle mobility prediction and ride request prediction to obtain an approximate-optimal vehicle dispatching strategy in an efficient way. The preliminary experiment based on the real-world data shows a significant improvement of the sensing coverage.

The effectiveness of rehabilitation in acute cruciate ligament (CL) rupture is dependent upon patient adherence to Home Exercise Programmes (HEPs) and self-efficacy. This paper presents the research, preliminary design stages and an early evaluation of a digital wearable system for monitoring, tracking, guiding and motivating users during HEP. The aim of the prototype is to support patients' rehabilitation programme by reducing the risk of re-injury during the process and motivate them to adhere to their HEPs by providing constructive feedback, encouraging understanding and thus promoting self-efficacy. The digital infrastructure is composed of three main parts, a physical product of two smart bracelets for sensing data from the patient's knee, a smartphone application for the user to interact with and a web-based service for collecting, storing, analysing and sharing data.

In1 order to minimize signal crosstalk, system noise and random noise interference and pseudo signals in the tactile sensing array to obtain high precision surface profile estimation, this paper explored the data processing technology of robot tactile sensing suit based on fuzzy cognition map. In the paper, it discussed the structure, the modeling process and steps of fuzzy cognition map, made the anatomy of the shape of an object's surface contour recognition, and researched on the algorithm description of cognition map of robot tactile suit and the implementation of tactile sensing data processing. The simulation experiment would be simple and clear to verify the force condition of each sensing point and the distribution profile of different pressure range and the distribution and size of pressure peak. The research results show that the proposed design solution is reasonable feasible for data processing of robot sensor suit.

Mobile Crowd Sensing (MCS) is a promising paradigm to collect large-scale network data globally. However, how to motivate people to collect and share data is a challenge. We believe the major reason why many MCS systems are not pervasive is because there are no incentives for people to use them. In this paper, we present SensingGO, a system which encourages people to keep sensing data by integrating incentive mechanisms. The sensed data are then transmitted to our backend server. The data we collected and the source code of SensingGO are open to anyone freely. We also demonstrate the analysis of real mobile data collected from SensingGO.

Momentary experiences of positive and negative emotionality--- also referred to as state affect---are core components of well-being and performance. The ability to unobtrusively monitor state affect could raise individuals' awareness of their mental health status and enable healthcare providers to deliver targeted, just-in-time mental health interventions. In this work, we investigate whether passively sensed smartphone data can be used to recognize individuals' state affect. Our exploratory analysis uses data generated from 220 participants in a two-week study, and our results indicate that fluctuations in participants' negative state affect are associated with various aspects of context including current physical state, geographic location, and time. We also test algorithms that predict participants' negative state affect given contextual features, comparing the impact of historical contextual features on prediction performance.

With the proliferation of sensor-rich mobile devices, crowd sensing has emerged as a new paradigm of collecting information from the physical world. However, the sensory data provided by the participating workers are usually not reliable. In order to identify truthful values from the crowd sensing data, the topic of truth discovery, whose goal is to estimate each worker's reliability and infer the underlying truths through weighted data aggregation, is widely studied. Since truth discovery incorporates workers' reliability into the aggregation procedure, it shows robustness to the data poisoning attacks, which are usually conducted by the malicious workers who aim to degrade the effectiveness of the crowd sensing systems through providing malicious sensory data. However, truth discovery is not perfect in all cases. In this paper, we study how to effectively conduct two types of data poisoning attacks, i.e., the availability attack and the target attack, against a crowd sensing system empowered with the truth discovery mechanism. We develop an optimal attack framework in which the attacker can not only maximize his attack utility but also disguise the introduced malicious workers as normal ones such that they cannot be detected easily. The desirable performance of the proposed framework is verified through extensive experiments conducted on a real-world crowd sensing system.

Backscatter WiFi offers a novel low-cost and low-energy solution for RFID tags to communicate with existing WiFi devices. State-of-the-art backscatter WiFi solutions have seldom explored advanced features in the latest WiFi standards, in particular, spatial multiplexing, which has been the cornerstone for 802.11n and beyond. In this paper, we present MOXcatter, a WiFi backscatter communication system that works with spatial streams using commodity radios, while keeping the ongoing data communication unaffected. In MOXcatter, a backscatter tag can embed its sensing data on ambient spatial-stream packets, and both the sensing data and the original packets can be decoded by commodity WiFi devices. We have built a MOXcatter prototype with FPGAs and commodity WiFi devices. The experiments show that MOXcatter achieves up to 50 Kbps throughput for a single stream and up to 1 Kbps for double streams with a communication range (tag-to-RX) up to 14 m. We discuss the tradeoffs therein and possible enhancements, and also showcase the applicability of our design through a sensor communication system.

This video presents a real-time end-to-end system which enables cameras to send personalized messages to people in a public area without knowing any addresses of their mobile phones. For facilitating this communication, we solve the problem of digitally associating people in the camera view with their smartphones without prior knowledge of the phones' IP/MAC addresses. The system doesn't need any dedicated devices and doesn't request people to wear digital tags. It utilizes users' motion patterns and leverages the diversity in motion features as the address for communication. The cameras broadcast a message to all the phones in the camera view using the target's motion features as the destination. Then a user's phone can locally compare the "motion address" of the packet against its own sensor data and will accept the packet if it's a "good" match. To protect the privacy of users' sensor data, we keep the users' personal sensing data on their phones instead of asking them to upload the data to server. Moreover, to prevent users' walking behavior from being revealed to public, we transform the raw motion features via principal component analysis (PCA) while maintaining their distinguishing power. On the whole, our system achieves 98%, 95%, 90%, 90%, 87% matching correctness for 2, 4, 6, 8 and 10 users respectively.

A major limitation of existing Semantic Web applications is the lack of automatic generation linked data for personal needs. Internet of Things (IoT) can provide automatic sensing data to improve this limitation. The study addresses this issue by defining a Semantic Internet of Things Framework (SIOTF), which is implemented on Hadoop-based cloud computing ecosystem to provide efficiency in dealing with a mass of sensing data. The SIOTF is composed of four modules: Internet of Things module, Naive Bayesian Classification module, Open Data Service module, and Semantic Web module. The proposed SIOTF is used to develop a Culture Sharing Cloud Platform (CSCP) that provides customized culture information for personnel needs. To demonstrate the feasibility of CSCP, the experimental results illustrate the efficiency and effectiveness of the proposed approach.

We describe Flux, a platform for dynamically reconfigurable data sensing using mobile devices, like smartphones or tablets. Periodic sensing tasks are programmed using the Flux Task Language and compiled onto abstract byte-code that is executed by a low-footprint virtual machine, guaranteeing by construction important runtime safety properties. For task dissemination, a Flux gateway performs on-the-fly injection of tasks on devices present in a geographical region, and sensing data is streamed back to the gateway that forwards it to a publish/subscribe broker. Live or archived streams are in turn fed by the broker to data processing clients. We implemented a prototype of Flux and used it to conduct a case-study experiment where the intensity of Wifi signal in our department is measured over a certain survey area, using smartphones and tablets carried by volunteers as they walked over the survey area.

Advances of IoT facilitates leverage of heterogeneous sensing data over the Internet, such as remote sensing data, traffic data and SNS data. Integrated analysis of IoT data is crucial part for urban emergency management in smart cities in order to predict various social events co-occurring with a natural disaster event. Discovering of spatiotemporal co-occurrence patterns is a task of integrated analysis of IoT data and has received a lot of attention. However, such spatiotemporal co-occurrence patterns can fail to capture local events that occur in limited regions and limited time intervals.In this paper, we consider the problem of mining spatiotemporal co-occurrence patterns from IoT sensing data, each of which is annotated with a valid spatial and temporal region. Our idea is to incorporate spatiotemporal clustering with the frequent itemset (pattern) discovery process to reduce spatiotemporal bias of event distributions and we repeat this process in greedy approach in order to capture patterns with difference scales. By this way, our algorithm improves accuracy of the frequent itemsets. We applied our method to discovery and prediction of traffic disaster events co-occurring with torrential rain events in Kansai area, Japan. Our experimental result shows 31% improvement of prediction performance on F-measure against a baseline.

In this paper, given the scenario of a mobile sensing system for air pollution monitoring, we aim at the cause and influence of delay effect on measurement and present a filter-based solution to calibrate the sensing data. We also validate the idea and solution by a real-data experiment. It indicates that the solution decreases deviation on spatial measurement and can be applied in mobile sensing systems to improve the sensing data quality.

We present a framework - AWSense - for eased sensing data collection from an Apple Watch wearable device. The framework eases the access, transmission and export of sensing data from the device. This data comprises: heart rate, raw acceleration, and computed device motion. In our demo, we present sample applications built on top of this framework to show its capabilities, of real-time presentation and recording of the sensing data.

Social sensing has received growing interest in a broad range of applications from business to health care. The potential benefits of modeling infectious disease spread through geo-tagged social sensing data has recently been demonstrated, yet it has not considered contagion events that can occur even when co-located individuals are no longer in physical contact, such as for capturing the dynamics of airborne diseases. In this study, we exploit the location updates made by 0.6 million users of the Momo social networking application to characterize airborne disease dynamics. Airborne diseases can transmit through infectious particles exhaled by the infected individuals. We introduce the concept of same-place different-time (SPDT) transmission to capture the persistent effect of airborne particles in their likelihood to spread a disease. Because the survival duration of these infectious particles is dependent on environmental conditions, we investigate through large-scale simulations the effects of three parameters on SPDT-based disease diffusion: the air exchange rate in the proximity of infected individuals, the infectivity decay rates of pathogen particles, and the infection probability of inhaled particles. Our results confirm a complex interplay between the underlying contact network dynamics and these parameters, and highlight the predictive potential of social sensing for epidemic outbreaks.

Telepath is a framework that supports communication-free offloading for wearable devices. With offline training, activity recognition tasks can be offloaded from the wearable to the user's phone, without transferring raw sensing data. The key observation is that when the user is carrying both devices, the sensing streams on the two devices are highly correlated. By exploiting the correlation, the phone can estimate the wearable's sensing data and emulate the watch. Our evaluations shows that with Telepath, the phone performs accurately on activity recognition tasks that are designed for smart watches, achieving on average 87% of the watch's accuracy while extending the watch's battery life by 2.1x.

Driven by the proliferation of sensor-rich mobile devices, crowd sensing has emerged as a new paradigm of gathering information about the physical world. In crowd sensing applications, user observations are usually unevenly distributed across the monitored entities, and this gives rise to two major challenges -- redundancy and sparsity. On one hand, multiple users may observe the same entity, and their observations are sometimes conflicting with each other due to the unreliable nature of human-carried sensors. On the other hand, crowd sensing data are usually very sparse, and there may exist considerable number of entities that never receive any observations from users. Some existing work studies these two challenges separately. However, we can gain great benefits by dealing with them jointly. In this paper, we develop an integrated framework to estimate the true values of entities from redundant and sparse data in crowd sensing applications. In this framework, we propose an effective algorithm to infer the "missing" observations for each entity, and aggregate both user-contributed and inferred observations to discover the true values of entities. We conduct extensive experiments on real-world crowd sensing systems to demonstrate the advantages of the proposed framework on correctly inferring entity truths from redundant and sparse data.

Daily toothbrushing is essential for maintaining oral health. However, there is very limited technology to monitor the effectiveness of toothbrushing at home. In this paper, a system is built to monitor the brushing quality on all 16 tooth surfaces using a manual toothbrush and an off-the-shelf wrist watch. The toothbrush is modified by attaching small magnets to the handle, so that its orientation and motion can be captured by the magnetic sensor in the wrist watch. The toothbrushing gestures are recognized based on inertial sensing data from the wrist watch. As the acoustic signal collected from the watch is correlated with the motion of toothbrushing stroke, acoustic sensing algorithm is designed to assist in recognition. User-specific toothbrushing order is also utilized to improve the surface recognition. In extensive experiments with 12 users over 3 weeks, our system successfully recognized toothbrushing gestures with an average precision of 85.6%.

Citizen science and communitysensing applications allow everyday citizens to collect data about the physical world to benefit science and society. Yet despite successes, current approaches are still limited by the number of domain-interested volunteers who are willing and able to contribute useful data. In this paper we introduce habitsourcing, an alternative approach that harnesses the habit-building practices of millions of people to collect environmental data. To support the design and development of habitsourcing apps, we present (1) interaction techniques and design principles for sensing through actuation, a method for acquiring sensing data from cued interactions; and (2) ExperienceKit, an iOS library that makes it easy for developers to build and test habitsourcing applications. In two experiments, we show that our two proof-of-concept apps, ZenWalk and Zombies Interactive, compare favorably to their non-data collecting counterparts, and that we can effectively extract environmental data using simple detection techniques.

In this paper, we report our efforts to recognize human contexts through the things sensorized by our sensor platform. As you know, many things wearing on your body have already become a sensor to understand our activities. Most famous things are watches and glasses. In the future, we suppose that unwearable things, such as cutlery, sport gear and instruments, would be also a sensor for recognizing our activities. Compared with our movements, the movement of unwearable things will have more variation. Also, suitable sensor and proper algorithm would be different according to the things. To enhance the senserization of various things, we have proposed a platform, called SenStick, that consists of ultra tiny all-in-one sensor board and sophisticated mobile app. If using it, every researcher can try embedding various sensor into various things and can focus on data analysis. We explain the possibility of smart experiment platform with introducing various sensorized things on the platform.

We analyze over 100 Gb of electric bicycle (e-bike) usage data collected through the University of Waterloo WeBike field trial. The WeBike fleet consists of 31 instrumented e-bikes used by University of Waterloo faculty, staff and students. We break down usage and battery charging habits by gender and by occupation, and we compare participants' initial estimates of how much they thought they would ride their e-bike with their actual riding histories. We also discuss data pre-processing challenges such as identifying trip start and end times from noisy and incomplete sensing data.

Players often have playfulness and motivation to play and make games on their own. By tapping into these instincts and guiding players in designing their own games to solve a problem, new models of collaboration between designers and players can be unlocked. In support of this, we present results from a five week case study with 25 children aged 8-10 where they designed their own games within a given context. This ongoing exploration resulted in three design themes: 1) Setting a common goal helps people design together, in this case using sensing data. 2) Focusing on spontaneity through autonomy. And 3) Supporting player mobility with the artefacts. We hope this ongoing exploration can be a starting point in aiding designers to guide players as designers.

In this work, an "analytical data model of mosquito vector" was developed to perform analytical computation to the character of the dengue vectors. Our goal is to investigate a way to understand how the temporal trend of collected dataset correlates with the incidence dengue as identified by national health authorities. Based upon the mosquito-vector big data collections, we investigate how changes in some specific variables such as rainfall, temperature, and humidity can dramatically affect the population of mosquito vectors, in order to provide early warnings of dengue outbreaks. Thus, our system will collectively collect online sensing data of the variables and store them in a database, in order to combine the historical big data as training datasets for analytical computation. Also, the developed model is able to merge the experimental datasets with current hot-topic information which is relevant to mosquito vectors obtained from data of social sensors (i.e. social messages). The experimental data show that our system is of great potentials in providing early warnings of dengue outbreaks.

Cognitive Radio Network (CRN) nodes called the Cognitive Radio (CR) nodes, sense the presence of a Primary User (PU) activity. The sensing reports from all the CR nodes are sent to a Fusion Centre (FC) which aggregates these reports and takes decision about the presence of the PU, based on some decision rules. Such a collaborative sensing mechanism forms the foundation of any centralised CRN. However, this approach can be misused by malicious Secondary Users (SUs) by carrying out Spectrum Sensing Data Falsification (SSDF) attacks. SSDF attacks which invade the CRN during spectrum sensing phase can affect the global decision of spectrum occupancy and thus, degrade the overall performance of a CRN. In this paper, a threshold-based detection strategy, CODES, is proposed for detection of SSDF attacks. Simulation results show successful detection of malicious SUs across a range of SSDF attacks.

In crowdsensing, appropriate rewards are always expected to compensate the participants for their consumptions of physical resources and involvements of manual efforts. While continuous low quality sensing data could do harm to the availability and preciseness of crowdsensing based services, few existing incentive mechanisms have ever addressed the issue of sensing data's quality. The design of quality based incentive mechanism is motivated by its potential to avoid inefficient sensing and unnecessary rewards. In this paper, we incorporate the consideration of data quality into the design of incentive mechanism for crowdsensing, and propose to pay the participants as how well they do, to motivate the rational participants to perform data sensing efficiently. This mechanism estimates the quality of sensing data, and offers each participant a reward based on her effective contribution. We also implement the mechanism and evaluate the improvements in terms of quality of service and profit of service provider. The evaluation results show that our mechanism achieves superior performance when compared to the uniform pricing scheme.

Supervisory control for multiple unmanned aerial vehicles (UAVs) operated by a single operator has been studied [1], The operator decides a plan by which he/she completes a mission such as surveillance and rescue. A supporting system for the operator is needed to select an optimal plan using a large number of data such as sensing data from the UAVs and a map from a cloud. On the other hand, in cloud computing, potentially unlimited computing and database resources can be utilized. In this paper, we apply a cyber-physical system approach to the design of a system for a Cloud-Assisted Sensing and Supervisory control (CLASS) of the multiple UAVs operated by a single operator. The cloud provides a list of plans from which the operator selects an optimal one, a user interface suitable for mission accomplishment, information such as the UAVs' data and a map to the operator, and commands to each UAV for an automatic navigation. Moreover, the cloud retrieves data used to complete the mission from other clouds.

An increasing popularity of smartphones affects interaction behaviors between people. The ubiquity of smartphones with embedded sensors can enable attractive sensing applications for collecting and reporting data in participatory sensing environment. However, time and monetary cost hinder people from providing sensing data using their smartphones. Smartphone users will probably not participate in sensing tasks without incentive in the real world. Therefore, we propose SenseUtil which provides participation-aware incentive for crowd sensing environment. The main objective of SenseUtil is to keep moderate payment. SenseUtil applies the concept of microeconomics, where demand and supply determine the value of sensed data. Several factors including sensing frequency, nearby sensing points and users' preference, affect the incentive which dynamically changes over time. We implemented a simulation to study the impact of participation-aware incentive mechanisms. The results demonstrate that historical data of participation activities help decrease payments moderately.

Emerging wearable devices provide a new opportunity for mobile context-aware applications to use continuous audio/video sensing data as primitive inputs. Due to the high-datarate and compute-intensive nature of the inputs, it is important to design frameworks and applications to be efficient. We present the GlimpseData framework to collect and analyze data for studying continuous high-datarate mobile perception. As a case study, we show that we can use low-powered sensors as a filter to avoid sensing and processing video for face detection. Our relatively simple mechanism avoids processing roughly 60% of video frames while missing only 10% of frames with faces in them.

For understanding human activities in urban public spaces, we have been developing a sensing platform for visualizing data derived from sensor devices in the spaces. The platform collects sensing data from sensor devices installed in public spaces, and provide a simple communication interface to access the sensor data for client systems. In this demo abstract, we describe a working prototype of a sensor data streaming service for visualizing urban public spaces. The streaming data includes the number of pedestrians, pedestrian flows, and sound pressures of the human activities in the sensing area. As an application of the streaming service, we will show a client system that visualizes the current pedestrian flows in the sensing area with very little delay. The flows are visualized like a view of live camera without pedestrians' appearances. The system has a search interface to retrieve the sensing data in a past period and also visualizes them.

In this paper we consider the use of a link-unreliable wireless sensor network for remote monitoring, where the monitoring center is geographically located far away from the region of the deployed sensor network. The sensing data is transferred to the monitoring center by the third party communication service, which incurs service cost. We first formulate a novel optimization problem of maximizing the network throughput with minimal service cost, which is shown to be NP-hard. We then develop approximation algorithms. We finally evaluate the performance of the proposed algorithms by simulations. Experimental results demonstrate that the solutions delivered by proposed algorithms are fractional to the optimum.

By interacting with external sensors, smartphones can gather high-fidelity data on the surrounding environment to develop various environment-aware, personalized applications. In this work we introduce the sensor virtualization module (SVM) which virtualizes external sensors so that smartphone applications can easily utilize a large number of sensing resources. Implemented on the Android platform, our SVM simplifies the management of external sensors by abstracting them as virtual sensors to provide the capability of resolving conflicting data requests from multiple applications and also mashing-up sensing data from different sensors to create new customized sensors. We envision our SVM to open the possibilities of designing novel personalized smartphone applications

Emerging mobile sensing applications are changing the characteristics of smartphone workloads. Whereas typical mobile applications run alone in the foreground interacting with users, sensing applications concurrently run in the background, providing unobtrusive monitoring services. Such concurrent sensing workloads raise a new challenge incurring severe resource contention among themselves and with other foreground applications. To address the challenge, we develop SymPhoney, a coordinated sensing flow execution engine to support concurrent sensing applications. As its key approach, we develop a novel sensing-flow-aware coordination. We first introduce the new concept of frame externalization i.e., to identify and externalize semantic structures embedded in otherwise flat sensing data streams. Leveraging the identified frame structures, SymPhoney develops frame-based coordination and scheduling mechanisms, which effectively coordinates the resource use of concurrent contending applications and maximize their utilities even under severe resource contention. We implemented several sensing applications on top of the SymPhoney engine and performed extensive experiments, showing effective coordination capability of SymPhoney.

It is very important to be able to determine humans' locations and their movement in order to a ubiquitous service environment. In this paper, we propose an algorithm that simultaneously estimates the number of humans and the movement path for each human in a room, using only the binary sensing data obtained from infrared sensors attached to the ceiling. Compared to other camera-based systems, the estimation results of our algorithm can be used directly in application systems without violating anyone's privacy. After describing the proposed algorithm in detail, we then describe our evaluation results.

For a few years, u-City[1] project in Korea have been based on USN(Ubiquitous Sensor Network)[2--3]. Thus, there are many studies to utilize the Sensor Network effectively. But, in order to utilize these results for u-City, it requires erasing complexity driven by heterogeneity of a wide range of sensors and support compatibility with a lot of components in u-City. Therefore, we have built the SensorBridge, as a message delivery system, to fill the gap between the heterogeneous devices and multi-users, allow the aptitudes for bidirectional connecting between devices and users, fetching real-time data, and compatible with a large range of devices. In this paper, we will describe about designing of SensorBridge, as well as the results in some real projects.

While the characterization of food consumption level has been extensively studied in nutrition and psychology research, advancements in passive smartphone sensing have not been fully utilized to complement mobile food diaries in characterizing food consumption levels. In this study, a new dataset regarding the holistic food consumption behavior of 84 college students in Mexico was collected using a mobile application combining passive smartphone sensing and self-reports. We show that factors such as sociability and activity types and levels have an association to food consumption levels. Finally, we define and assess a novel ubicomp task, by using machine learning techniques to infer self-perceived food consumption level (eating as usual, overeating, undereating) with an accuracy of 87.81% in a 3-class classification task by using passive smartphone sensing and self-report data. Furthermore, we show that an accuracy of 83.49% can be achieved for the same classification task by using only smartphone sensing data and time of eating, which is an encouraging step towards building context-aware mobile food diaries and making food diary based apps less tedious for users.