High-resolution wide-angle fisheye images are becoming more and more important for robotics applications such as autonomous driving. However, using ordinary convolutional neural networks or vision transformers on this data is problematic due to projection and distortion losses introduced when projecting to a rectangular grid on the plane. We introduce the HEAL-SWIN transformer, which combines the highly uniform Hierarchi-cal Equal Area iso-Latitude Pixelation (HEALPix) grid used in astrophysics and cosmology with the Hierarchical Shifted-Window (SWIN) transformer to yield an efficient and flexible model capable of training on high-resolution, distortion-free spherical data. In HEAL-SWIN, the nested structure of the HEALPix grid is used to perform the patching and windowing operations of the SWIN transformer, enabling the network to process spherical representations with minimal computational overhead. We demonstrate the superior performance of our model on both synthetic and real automotive datasets, as well as a selection of other image datasets, for semantic segmentation, depth regression and classification tasks. Our code is publicly available11https://github.com/JanEGerken/HEAL-SWIN.

We analyze the role of rotational equivariance in convolutional neural networks (CNNs) applied to spherical images. We compare the performance of the group equivariant networks known as S2CNNs and standard non-equivariant CNNs trained with an increasing amount of data augmentation. The chosen architectures can be considered baseline references for the respective design paradigms. Our models are trained and evaluated on single or multiple items from the MNIST- or FashionMNIST dataset projected onto the sphere. For the task of image classification, which is inherently rotationally invariant, we find that by considerably increasing the amount of data augmentation and the size of the networks, it is possible for the standard CNNs to reach at least the same performance as the equivariant network. In contrast, for the inherently equivariant task of semantic segmentation, the non-equivariant networks are consistently outperformed by the equivariant networks with significantly fewer parameters. We also analyze and compare the inference latency and training times of the different networks, enabling detailed tradeoff considerations between equivariant architectures and data augmentation for practical problems.

Tackling current challenges in production automation requires the involvement of new concepts like Internet of Things, System-of-Systems and local automation clouds. The objective of this paper is to address the actual process of defining a cloud based automation system. More specifically the design, engineering, operation and maintenance of an automation system must be captured and managed between all stakeholders involved. This is critical to create the expected benefits from the local automation cloud approach. This paper addresses the capability of capturing plant designs and coordinating information exchange based on the captured architecture. For this purpose an architectural component - Plant Description - is proposed to be used in the Arrowhead Framework, based on already existing plant automation standards. An overview of methodologies on how it can interact with the Arrowhead Framework's Orchestration process describes the usefulness in managing large-scale automation systems. A qualitative evaluation for one of the proposed approaches is also described in a water control use case that can be found both in process and building automation.

Current challenges in production automation requires the involvement of new technologies like Internet of Things (IoT), Systems of Systems and local automation clouds. The objective of this paper is to address the actual process of defining a cloud based automation system. The objective of this paper is to address one of the challenges involved in establishing and managing a cloud based automation system. Three key capabilities have been identified as required to create the expected benefits of local automation clouds; 1) capturing of plant design 2) capturing and distributing configuration and deployment information 3) coordinating information exchange. This paper addresses the capturing and distribution of configuration and deployment information. For this purpose a system service is proposed, the Configuration Store, following the principles of the Arrowhead Framework. The service is accompanied by a deployment methodology and a bootstrapping procedure. These are discussed for several types of automation technology, e.g. controllers, sensors, actuators. A qualitative evaluation of the proposed approach is made for four use cases; Building automation, Manufacturing automation, Process automation and IoT devices. Concluding the usability for large-scale deployment and configuration of Industrial Internet of Things.

The emergence and deployment of connected devices in many domains of application (e.g. industrial production, buildings and facilities, urban environment, etc.) have resulted in the need to achieve integration of multiple and more complex systems. This new environment is stressing the intrinsic limits imposed by monolithic standards, data models and integration methods that focus on specific domains of application, types of systems, or specific aspects of a system. This paper describes the Plant Description Service developed as part of the Arrowhead Interoperability framework (EU ECSEL funded project). The manuscript contains a description of the abstract system descriptive model based on which the Plant Description service was implemented, and describes how the service can be used to achieve integration of several industry standards and data models. One use case and one case study is provided that illustrates how the service was practically implemented to support engineering scenarios in the domain of industrial production. The paper concludes with a critical review of the approach and suggestion for future work and developments.

Supervision and control systems are being deployed in industrial environments such as mining plants and manufacturing facilities to ensure a continuous and effective production at a minimum cost. Such systems monitor a whole range of devices and collect their data for several purposes like maintenance and control operations.

The procedure of migrating SCADA and DCS functionality of the ISA-95 process automation architecture to a Service based automation architecture is discussed. Challenges in such migration are discussed and defined. From here the necessary migration technology and procedures are proposed. The critical migration technology is based on the mediator concept. The migration procedure is based on a functionality perspective and comprises four steps: initiation, configuration, data processing and control execution. Its argued that these steps are necessary for the successful migration of DCS and SCADA functionality in to the automation cloud.