Mobile Robot Cooperation with Infrastructure For Surveillance: Towards Cloud Robotics Hadi Aliakbarpour, Paulo Freitas, Joao Quintas, Christiana Tsiourti, and Jorge Dias Institute of Systems and Robotics, University of Coimbra. {hadi,pfreitas, jquintas}@isr.uc.pt, ctsiourti@ipn.pt, jorge@isr.uc.pt Abstract. This paper proposes a cooperative framework among a mobile robot and an infrastructure for the sake of surveillance and identification. The infras- tructure is comprised of a network of cameras, inertial sensors, a facial identi- fication system and a GPU-enabled server. The cooperation among the mobile robot and infrastructure is service-based. The scene is observed by camera sen- sors. There is a mobile robot freely moving within the scene. For some security reasons the system gets interested in identification of a person inside the scene. The infrastructure provides three services to the mobile robot including positions of the person and robot as well as 3D information of the person. In the other sides, the mobile robot provides the facial (portrait) image of a requested person, as a service. Using these services the infrastructure can identify the person. The cameras in the infrastructure and also the camera on the mobile robot are rigidly coupled with an inertial sensor (IS). The inertial data provided by IS is used in two ways: definition of a fusion-based virtual camera and also virtual planes to register 3D data. Taking advantage of the defined virtual camera, a lo- calization method is proposed to obtain the position of the robot with respect to the infrastructure. Experimental results are provided to demonstrate the feasibil- ity and effectiveness of the proposed framework for the purpose of being used in cloud robotics. Keywords: Mobile robot, infrastructure, cloud robotics, inertial sensor (IS), camera network, sensor fusion, localization. 1 Introduction In this paper a cooperative framework among a mobile agent and an infrastructure is proposed which is towards the cloud robotics. Several research groups are exploring the idea of robots that rely on cloud computing infrastructure to access vast amounts of processing power and data. This approach, which some are calling “cloud robotics,” would allow robots to off-load compute-intensive tasks like image processing and voice recognition and even download new skills instantly. For conventional robots, every task — moving a foot, grasping things, recognizing a face—requires a significant amount of processing and pre-programmed information. As a result, sophisticated systems such as humanoid robots need to carry powerful computers and large batteries to power them. According to James Kuffner, from Carnegie Mellon University, cloud-enabled robots