Digital signage is widely used in various fields such as transport systems, trading outlets, entertainment, and others, to display information in the form of images, videos, and text. The reliability of these resources, availability of required services and security measures play a key role in the adoption of such systems. Efficient management of the digital signage system is a challenging task to the service providers. There could be many reasons that lead to the malfunctioning of this system such as faulty displays, network, hardware or software failures that are quite repetitive. The traditional process of recovering from such failures often involves tedious and cumbersome diagnosis. In many cases, technicians need to physically visit the site, thereby increasing the maintenance costs and the recovery time. In this thesis, we propose a solution that monitors, diagnoses and recovers from known failures by connecting the displays to a cloud. A cloud-based remote and autonomous server configures the content of remote displays and updates them dynamically. Each display tracks the running process and sends the trace and system logs to the server periodically. These logs, stored at the server optimally using a customized log management module, are analysed for failures. In addition, the displays incorporate self-recovery procedures to deal with failures, when they are unable to create connection to the cloud. The proposed solution is implemented on a Linux system and evaluated by deploying the server on the Amazon Web Service (AWS) cloud. The main result of the thesis is a collection of techniques for resolving the display system failures remotely.