How many people are needed for a crowdsensing campaign?

Abstract

Environmental scientists, city planners, park administrators, and similar others, need to base their decisions on up-to-date, accurate and high resolution measurement data. However, acquiring such data by hand or by deploying sensor networks is expensive and time consuming. People without proper scientific training want to take part in the scientific process and assist with the needs of their respective communities. Crowdsensing is a technology-driven area where ICT platforms are being developed which permit anyone to participate in processes that help expand our understanding and improve our surroundings. Crowdsensing refers to the process in which crowds (large number of people) measure specific features and share the resulting data or send it to a central location in which it can be used by the people that need it. There have been many crowdsensing applications. Since the increase in popularity of smartphones, which are now ubiquitous, crowdsensing is more popular than ever. Smartphones have all the basic requirements for the deployment of a crowdsensing application: a vast number of sensors, WiFi and LTE connectivity to transfer the data, and powerful, easy to program processors, as well as Bluetooth which permits the addition of specialized sensors. A lot of people can now participate to crowdsensing but we still don't know how many are needed for a crowdsensing campaign to be successful. In order to answer this question, we built a simulator that mimics the characteristics of a crowdsensing campaign. We showcase three different scenarios in which we estimate the required number of participants and offer a discussion on the plausibility of having that many participants by taking into account factors such as accessibility to the area of interest (the area from which the measurements are needed).