Ubiquitous sensors, connected devices, smart objects, networking advances, and diverse data sets are the driving force towards a smart world within which computational intelligence is embedded in the physical environment to provide trustworthy, personalized, and adaptive services to people. This ubiquitous intelligence (UI) changes the computing landscape by enabling new breeds of applications and systems previously impossible. For example, by coupling everyday objects with intelligence (i.e.,sensing, computing, and reasoning capabilities), many tasks can now be simplified and automated. As a result, our living and working environments become smarter, more comfortable, and more efficient.

 

Crowdsensing become an important part of u biquitous computing, and it has found pervasive use in a variety of applications ranging from environment/residential monitoring, intelligent transportation and traffic planning, urban dynamic ssensing, public health and safety, to location based services. The need of understanding urban, society and environment dynamics to enable large-scale ubiquitous computing applications leads to a number of challenges on crowd sensing technology, such as:

 

• Task allocation. Considering the large population of mobile nodes, a sensing task must identify which node(s) may accept a task. Meanwhile, a set of criteria should be considered in filtering irrelevant nodes, such as the specific action of a required region (e.g., a particular street) and time window, acceptance conditions, device capabilities, and termination conditions.

• Team formation. Interactions among volunteers are necessary during the sensing process, but absent in most existing crowd sensing systems. Therefore, grouping users and facilitating the interaction among them should be a challenge of CSUI.

• Incentive mechanisms. To sense and process the desired data, participating devices may incur energy and monetary costs, or even explicit efforts from their owners.Without strong incentives, individuals may not be willing to participate in the sensing task with cost of their own limited resources. Therefore, a successful crowd sensing system must have an appropriate incentive mechanism to recruit,engage, and retain its participants

• DataRedundancy, Quality, and Inconsistency. In crowd sensing, there can be multiple participants involved in the same sensing activity, providing data with various quality. Therefore, CSUI would raise data redundancy and inconsistency issues.

• Ubiquitous intelligence extraction. With the increase in the large-scale, inter linked data collected from crowds, advanced techniques on mining, association, aggregation,and semantic fusion of the crowdsourced cross space and heterogenous data will become more and more important.

• Trust, security andprivacy issues. Malicious users may deliberately pollute crowdsourced data fortheir own benefits. Therefore, trust maintenance and abnormal detection methods should be developed to determine the trustworthiness and quality of collecte ddata. Meanwhile, to motivate user participation, a crowd sensing system must be capable of effectively protecting the privacy of participants while allowing their devices to reliably contribute high-quality data to these large-scale applications. 

To address the core technical challenge inemerging crowd sensing and ubiquitous intelligence technologies, the CSUI 2015 workshop aims to complement the main UIC2015 conference by setting the focus of the workshop on thefollowing topics:

 

• Crowdsensing frameworks and wireless localization technologies

• Incentive models and mechanisms

• Crowdsourced data processing and mining approaches and algorithms

• Crowdsensing systems and applications

We believe that these crowd sensing technologies together with the extracted ubiquitous intelligence will be able to facilitate more and more services and applications that are previously impossible.