Real-Time Task Assignment in Hyperlocal Spatial Crowdsourcing under Budget Constraints Hien To, Liyue Fan, Luan Tran, Cyrus Shahabi University of Southern California Los Angeles, CA Email: {hto,liyuefan,luantran,shahabi}@usc.edu Abstract—Spatial Crowdsourcing (SC) is a novel platform that engages individuals in the act of collecting various types of spatial data. This method of data collection can significantly reduce cost and turnover time, and is particularly useful in environmental sensing, where traditional means fail to provide fine-grained field data. In this study, we introduce hyperlocal spatial crowdsourcing, where all workers who are located within the spatiotemporal vicinity of a task are eligible to perform the task, e.g., reporting the precipitation level at their area and time. In this setting, there is often a budget constraint, either for every time period or for the entire campaign, on the number of workers to activate to perform tasks. The challenge is thus to maximize the number of assigned tasks under the budget constraint, despite the dynamic arrivals of workers and tasks as well as their co- location relationship. We study two problem variants in this paper: budget is constrained for every timestamp, i.e. f ixed, and budget is constrained for the entire campaign, i.e. dynamic. For each variant, we study the complexity of its offline version and then propose several heuristics for the online version which exploit the spatial and temporal knowledge acquired over time. Extensive experiments with real-world and synthetic datasets show the effectiveness and efficiency of our proposed solutions. Index Terms—Crowdsourcing, Spatial Crowdsourcing, Mobile Crowdsensing, Online Task Assignment, Budget Constraints. I. I NTRODUCTION With the ubiquity of smart phones and the improvements of wireless network bandwidth, every person with a mobile phone can now act as a multimodal sensor collecting and sharing various types of high-fidelity spatiotemporal data instanta- neously. In particular, crowdsourcing for weather information has become popular. With a few recent apps, such as mPING 1 and WeatherSignal 2 , individual users can report weather condi- tions, air pollutions, noise levels, etc. In fact, Dorminey in [6] regards crowdsourcing as the future of weather forecasting. Through our collaboration with the Center for Hydrome- teorology and Remote Sensing (CHRS) 3 at the University of California, Irvine, we have developed a mobile app, iRain 4 , to perform spatial crowdsourcing for precipitation informa- tion. Unlike other weather crowdsourcing apps, iRain allows CHRS researchers to request rainfall information at specific locations and times where their global satellite precipitation 1 http://mping.nssl.noaa.gov/ 2 http://weathersignal.com 3 http://chrs.web.uci.edu/ 4 https://play.google.com/store/apps/details?id=irain.app estimation technologies 5 fail to provide real-time, fine-grained data. Individual iRain users around those locations can re- spond to those requests by reporting rainfall observations, e.g., heavy/medium/light/none, and they can also issue rainfall information requests by “subscribing” to regions of interest. In general, spatial crowdsourcing (SC) [10] offers an effec- tive data collection platform where data requesters can create spatial tasks dynamically and workers are assigned to tasks based on their locations. Figure 1 depicts the architecture of iRain. A requester issues a set of rainfall observation tasks to the SC-server (Step 1) where each task corresponds to a specific geographical extent, e.g., a circle. The workers continuously update their locations to the SC-server when they become available for performing tasks (Step 0). Subsequently, the SC-server crowdsources the tasks among the workers in the task regions and sends the collected data back to the requester (Steps 2, 3). SC-Server Workers Requesters 2. Selected workers 1. Hyper-local SC tasks Push Notification Service 3. Task notifications 0. Report locations A, +5 mins C B D E Fig. 1: Hyperlocal spatial crowdsourcing framework. One major difference from existing SC paradigms [11], [10], [8], [17], [18], [20] is that workers in our paradigm do not need to travel to the exact task locations, e.g., to the centers of the circular regions, and are eligible to perform tasks as long as they are in close spatiotemporal vicinity of the tasks, e.g., enclosed in the circular regions 6 . We denote this new paradigm as Hyperlocal Spatial Crowdsourcing. The reason is twofold. Without requiring the workers travel physically, our paradigm lowers the threshold for worker participation and will poten- tially yield faster response. Furthermore, the requested data, e.g., rainfall or temperature, exhibits spatiotemporal continuity 5 http://hydis.eng.uci.edu/gwadi/ 6 Tasks that require workers to physically travel to task locations, e.g., taking a picture of an event, are not considered in our problem setting. XXX-X-XXXX$XX.00 c 2016 IEEE