This paper highlights cloud computing as one of the principal building blocks of a smart factory, providing a huge data storage space and a highly scalable computational capacity. The cloud computing system used in a smart factory should be time-predictable to be able to satisfy hard real-time requirements of various applications existing in manufacturing systems. Interleaving an intermediate computing layer-called fog-between the factory and the cloud data center is a promising solution to deal with latency requirements of hard real-time applications. In this paper, a time-predictable cloud framework is proposed which is able to satisfy end-to-end latency requirements in a smart factory. To propose such an industrial cloud framework, we not only use existing real-time technologies such as Industrial Ethernet and the Real-time XEN hypervisor, but we also discuss unaddressed challenges. Among the unaddressed challenges, the partitioning of a given workload between the fog and the cloud is targeted. Addressing the partitioning problem not only provides a resource provisioning mechanism, but it also gives us a prominent design decision specifying how much computing resource is required to develop the fog platform, and how large should the minimum communication bandwidth be between the fog and the cloud data center.

Internet of Things (IoT), one of the key elementsof a smart factory, is dubbed as Industrial IoT (IIoT). Softwaredefined networking is a technique that benefits network managementin IIoT applications by providing network reconfigurability.In this way, controllers are integrated within the networkto advertise routing rules dynamically based on network andlink changes. We consider controllers within Wireless SensorNetworks (WSNs) for IIoT applications in such a way to providereliability and timeliness. Network reliability is addressed for thecase of node failure by considering multiple sinks and multiplecontrollers. Real-time requirements are implicitly applied bylimiting the number of hops (maximum path-length) betweensensors and sinks/controllers, and by confining the maximumworkload on each sink/controller. Deployment planning of sinksshould ensure that when a sink or controller fails, the networkis still connected. In this paper, we target the challenge ofplacement of multiple sinks and controllers, while ensuring thateach sensor node is covered by multiple sinks (k sinks) andmultiple controllers (k' controllers). We evaluate the proposedalgorithm using the benchmark GRASP-MSP through extensiveexperiments, and show that our approach outperforms thebenchmark by lowering the total deployment cost by up to24%. The reduction of the total deployment cost is fulfilled notonly as the result of decreasing the number of required sinksand controllers but also selecting cost-effective sinks/controllersamong all candidate sinks/controllers.

The increasing functionality and complexity of automotive applications requires not only the use of more powerful hardware, e.g., multi-core processors, but also efficient methods and tools to support design decisions. Component-based software engineering proved to be a promising solution for managing software complexity and allowing for reuse. However, there are several challenges inherent in the intersection of resource efficiency and predictability of multi-core processors when it comes to running component-based embedded software. In this paper, we present a software design framework addressing these challenges. The framework includes both mapping of software components onto executable tasks, and the partitioning of the generated task set onto the cores of a multi-core processor. This paper aims at enhancing resource efficiency by optimizing the software design with respect to: 1) the inter-software-components communication cost, 2) the cost of synchronization among dependent transactions of software components, and 3) the interaction of software components with the basic software services. An engine management system, one of the most complex automotive sub-systems, is considered as a use case, and the experimental results show a reduction of up to 11.2% total CPU usage on aquad-core processor, in comparison with the common framework in the literature. 

Cloud computing is receiving an increasing attention when it comes to providing a wide range of cost-effective services. In this context, energy consumption of communication and computing resources contribute to a major portion of the cost of services. On the other hand, growing energy consumption not only results in a higher operational cost, but it also causes negative environmental impacts. A large number of cloud applications in, e.g., telecommunication, multimedia, and video gaming, have real-time requirements. A cloud computing system hosting such applications, that requires a strict timing guarantee for its provided services, is denoted a Real-Time Cloud (RTC). Minimizing energy consumption in a RTC is a complicated task as common methods that are used for decreasing energy consumption can potentially lead to timing violations. In this paper, we present an online energy-aware resource provisioning framework to reduce the deadline miss ratio for real-time cloud services. The proposed provisioning framework not only considers the energy consumption of servers but it also takes the energy consumption of the communication network into account, to provide a holistic solution. An extensive range of simulation results, based on real data, show a noticeable improvement regarding energy consumption while keeping the number of timing violations less than 1% in average.