• mobile sensing
• pattern classification
• urban sensing

The continuous and growing spread of last generation mobile devices has made possible to consider the end user not only as the simple as a data consumer but also as a data producer. The contemporary smartphones offer a set of sensors that produce the measurements of environmental phenomenons: from those measurements, numerous information useful to a new class of environment-aware applications can be extracted. Such applications are the basis for urban sensing projects that analyses a series of urban issues relying on data collected from sensors. The work is about designing, implementing and testing a software system that identifies traffic conditions on streets starting from GPS location data collected from on-road vehicles. Identification is based upon specific traffic patterns built using a set of features that distinguish the different types of traffic conditions. Testing results show that higher than 90% identification accuracy can be achieved using this method.