Activity classification and user interface design for a crowdsourcing urban simulation platform using mobile devices

Publication Type:

Conference Paper


PED - Pedestrian and Evacuation Dynamics (2012)


crowdsourcing urban simulation; mobile devices


The research area of urban simulation methods has grown notably in recent decades. Most of the research topics that concern urban simulation have concentrated on defining the complexities of urban environments with certain rules and algorithms. However, cities are getting more complex and changes to them are being made at greater speed [2]. Therefore, current urban simulation modeling approaches based on rules and protocols are still struggling to reduce the gap between the virtual simulation environment and the real cities, since the behavior of citizens is frequently unpredictable and continuously adapting.
In this context, research is necessary to develop more fundamental simulation methods that can handle these complexities and changes, leading to new design decision support systems [3]. Therefore, this research was motivated with the following questions: What is the origin of the complexities and transformations of the urban environment? How can we approach the origin to deal with the urban complexities and transformations?
To answer these questions, we hypothesize that the diverse human behavior are the origin of the issues that result from all of the complexities and changes of the cities.
In this paper, we introduce the idea of a crowdsourcing urban simulation platform using smartphones [1, 4]. Such development requires research to be conducted in numerous disciplines: social sensing, urban sustainability, mobile network, behavior pattern analysis, and social network services. It is based on the detection and classification of activity patterns (currently mainly focusing on traffic information), and on state-of-the-art interactive user interfaces in order to let user utilize the application easily. On the basis of these cutting-edge technologies, this research aims at the design and implementation of a practical participatory urban sustainability simulation platform.
As a result, we propose a participatory simulation environment that feeds sensed human behavior into an urban simulator, and thereby includes urban complexities and dynamics simultaneously. This research pursues to collect urban behavior data through the smartphone and suggests possible user benefits for their data sharing.
Therefore this research has following objectives:
- Establish sensing methods for mass data collection using smartphones
The first step of this research was to set up the basis for mobile crowdsourcing. This includes defining what kinds of data should be collected, how the data needs to be transformed to be useful for urban simulation, and how information can be fed back to the user.
- Classify the user’s activity, focusing on transportation mode
We have developed a prototype mobile phone application that implements a novel transportation mode detection algorithm. The mobile phone application runs in the background and continuously collects data from the built-in acceleration and network location sensors. The collected data is analyzed by the transportation mode detection algorithm and automatically partitioned into activity segments. A key observation of our work is that walking activity can be robustly detected in the data stream. Therefore it is used as a separator for partitioning the data stream into other activity segments. Each vehicle activity segment is then sub-classified according to the type of taken vehicle. Our approach yields high accuracy at a low sampling interval and does not require GPS data. Therefore, device power consumption is minimized.
- Design and implement an effective user interface for contributing and sharing
In order to make crowdsourced data collection effective, mass participation is essential. Therefore it is essential to attract users and to keep them interested in an application. Thus, we will introduce the possible alternatives; show how to effectively deliver and visualize the users’ information, how to give benefits to users in return for their contributions, and suggest what kind of social interactions can be implemented to induce more participation.
Our goal is to enable people to share urban information at any time through the crowdsourcing urban simulation platform [5]. The information will be returned to the citizens to support their sustainability-aware life. The simulation platform also gives a chance not only to compare each other’s levels of sustainability, but also to give self-satisfaction through an altruistic contribution for a sustainable future. Thus, people shall utilize the simulator in order to predict their individual or cities’ future sustainability. Meanwhile, the user data will be collected and delivered to the central server in order to analyze the urban sustainability.
We present the methods collecting urban data using smartphone and activity classification from the acceleration and location data, and further, the user interface and user benefit are suggested to give a motivate for voluntary participation. Consiquently, we can measure the urban sustainability based on a real human interaction, and compare individuals as well as cities. The whole process of this research is presented as a new paradigm of an urban simulator that reflects the urban complexities and the inconstant human mind changes.