Article Preview
Top1. Introduction
The rapid growth in the population density in urban cities demands tolerable provision of services and infrastructure, in order to meet the needs of city inhabitants. Thus, the raise in the request for embedded devices, such as sensors, actuators, and smartphones, etc., which is providing a great business potential towards the new era of Internet of Things (IoT); in which all the devices are capable of interconnecting and communicating with each other over the Internet. Therefore, the Internet technologies provide a way towards integrating and sharing a common communication medium for collaborative urban systems (Rathore et al., 2016).
The mobile sensing revolution is coming of age, and we will soon see these systems in everyday use. This progress is accelerated by the development of smartphones as a viable sensing platform. Today, most mobile phones include various sensors, such as GPS, accelerometers, microphones, and cameras. Classification models can exploit such data to allow, for instance, a mobile phone to understand our actions and environment. These models are driving key mobile application domains including mobile heath and green energy awareness, to name a few. However, significant challenges exist in the real-world with respect to human activity modeling. For example, a key obstacle is the differences in contextual conditions and user characteristics encountered in large-scale mobile sensing systems. This leads to the discriminative features in sensor data, used by classifiers to recognize different human activities, varying from user to user.
Up-to-date the worldwide are becoming increasingly more urbanized, around the 50 percent of the global population are living in only a smaller percent of the Earth’s surface, maybe it only represents the one percent. Probably, this urbanization will grow at least 70 percent for the future years. Thus, cities have evolved into one of the most impressive and complex artifacts. The increasing development of ubiquitous technologies are producing a wealth of information, reflecting different features of our lives over the last decade. Digital traces emanating through our daily interactions with pervasive computing devices are valuable sources of data for capturing the pulse of the city in an astonishing degree of temporal and spatial details.
These raw data can be used to develop data analytical tools in order to improve the understanding of urban dynamics and consequently make urban systems more efficient. The increasing use of ubiquitous devices, including mobile phones and GPS navigation systems, creates a new sensing theory in which humans serve as distributed sensors in the city that is addressed to measure the urban dynamics. Consequently, smart cities are intended to improve public services. They are seeking ways to reduce inefficiencies and to be more agile in responding to citizens’ needs.
The fact as technology becomes truly pervasive implies to proceed from single-user or single-system to large-scale heterogeneous systems, involving many devices and individuals collaborating over different spatial and temporal scales. The technological drivers that could facilitates such a perspective change are well known:
- •
The spread of diverse sensing capabilities to every-day mobile devices. Nowadays smartphones have GPS, cameras, microphones, acceleration sensors, gyroscopes, magnetometers, and light sensors. In addition, Bluetooth and Wi-Fi scans can be used to infer the presence of people and infrastructure. Many manufacturers are also considering the inclusion of environmental sensors such as temperature or air pressure.
- •
Pervasiveness of mobile Internet connectivity. Most smartphone users today have constant Internet connectivity, so uploading data is possible almost everywhere and at any time, considering restrictions related to traffic volume.
- •
Confluence of mobile computing and social networking that increasingly includes automatically derived context information. Some examples include services, such as Foursquare Facebook, and LinkedIn that make a user’s location visible online and fitness-monitoring applications that support online data uploads for community-based competitions.