Service-oriented Architecture supports software to be composed from services dynamically. Selecting and composing appropriate services according to business process, policies and non-functional constraints is an essential challenge. This paper proposes a method for automatic selection of the most relevant service for composition based on non-functional properties and the user’s context. In doing this we also propose a method of obtaining and evaluating non-functional aspects.

The deluge of intelligent objects that are providing continuous access to data and services on one hand and the demand of developers and consumers to handle these data on the other hand require us to think about new communication paradigms and middleware. In hyper-scale systems, such as in the Internet of Things, large scale sensor networks or even mobile networks, one emerging requirement is to process, procure, and provide information with almost zero latency. This work is introducing new concepts for a middleware to enable fast communication by limiting information flow with filtering concepts using policy obligations and combining data processing techniques adopted from complex event processing.

The deluge of intelligent objects that are providing continuous access to data and services on one hand and the demand of developers and consumers to handle these data on the other hand require us to think about new communication paradigms and middleware. Based on requirements collected from scenarios from connected car, social networks, and factory of the future this thesis is developing new concepts for fast data processing for hyper-scale systems. In hyperscale systems, such as in the Internet of Things, one emerging requirement is to process, procure, and provide information with almost zero latency. This thesis is introducing new concepts for a middleware to enable fast communication by limiting information flow with filtering concepts using event policy obligations and combining data processing techniques adopted from complex event processing. Fast data processing has to deal with continuous data streams of events, providing a set of operators to manipulate, aggregate, and correlate data. This processing logic needs to be distributed. Distribution helps us to scale on one hand in terms of numbers of data sources (e.g. phones, cars, sensors) and on the other hand to parallelise processing in terms of grouping and partitions (e.g. regional). In our solution, event policies are injected as close as possible to the place where the data is born to optimise traffic. Filters, aggregations and rules help to process the data accordingly. Finally, communication paradigms or interaction patterns support mediation between classical service based request-response interaction and event-based data exchange. This all together builds a middleware enabling fast data processing for hyper-scale systems.

In the last years Service-Oriented Architectures have been extensively used in order to enable seamless interaction and integration among the various heterogeneous systems and devices found in modern factories. The emerging Industrial Automation Systems are increasingly utilizing them. In the cloud-based vision of IMC-AESOP such technologies take an even more key role as they empower the backbone of the new concepts and approaches under development. Here we report about the investigations and assessments performed to find answers for some of the major questions that arise as key when technologies have to be selected and used in an industrial context utilizing Service-Oriented Architecture (SOA) based distributed large scale Process Monitoring and Control system. Aspects of integration, real-timeness, distributeness, event-based interaction, service-enablement etc. are approached from different angles and some of the promising technologies are analysed and assessed. François Jammes, Bernard Bony, Philippe Nappey Schneider Electric, France e-mail: francois2.jammes@schneider-electric.com,bernard.bony@ schneider-electric.com,philippe.nappey@schneider-electric.com Stamatis Karnouskos SAP, Germany e-mail: stamatis.karnouskos@sap.com Armando W. Colombo Schneider Electric & University of Applied Sciences Emden/Leer, Germany e-mail: armando. colombo@schneider-electric.com,awcolombo@technik-emden.de Jerker Delsing, Jens Eliasson, Rumen Kyusakov Luleå University of Technology, Sweden, e-mail: jerker.delsing@ltu.se,jens.eliasson@ltu.se, rumen.kyusakov@ltu.se Petr Stluka Honeywell, Czech Republic e-mail: petr.stluka@honeywell.com Marcel Tilly Microsoft, Germany, e-mail: marcel.tilly@microsoft.com Thomas Bangemann ifak, Germany e-mail: thomas.bangemann@ifak.eu

The Internet of Things, large scale sensor networks or even in social media, are now well established and their use is growing daily. Usage scenarios in these fields highlight the requirement to process, procure, and provide information with almost zero latency. This work is introducing new concepts for enabling fast communication by limiting information flow through filtering concepts combined with data processing techniques adopted from complex event processing. Specifically we introduce a novel mediation services architecture using filter policies to reduce latency. The filter policies define when and what data services need to provide to the mediator and thus save on bandwidth. The filter policies describe temporal conditions between two events removing the need to keep a complete history while still allowing temporal reasoning. Promising experimental results highlight the advantages to be gained from the approach.

Service-oriented Architecture supports software to be composed from services dynamically. Selecting and composing appropriate services according to business process, policies and non-functional constraints is an essential challenge. This paper proposes a method for automatic selection of the most relevant service for composition based on non-functional properties and the user’s context. In doing this we also propose a method of obtaining and evaluating non-functional aspects.

Connected devices are expected to grow to 50 billion in 2020. Through our industrial partners and their use cases, we validated the importance of inflight data processing to produce results with low latency, in particular local and global data analytics capabilities. In order to cope with the scalability challenges posed by distributed streaming analytics scenarios, we propose two new technologies: (1) JStreams, a low footprint and efficient JavaScript complex event processing engine supporting local analytics on heterogeneous devices and (2) DiAlM, a distributed analytics management service that leverages cloud-edge evolving topologies. In the demonstration, based on a real manufacturing use case, we walk through a situation where operators supervise manufacturing equipment through global analytics, and drill down into alarm cases on the factory floor by locally inspecting the data generated by the manufacturing equipment.

Nowadays businesses as well as the Web require information to be available in real-time in order to reply to requests, make effective decisions and generally remain competitive. This in turn requires data to be processed in realtime. In general in service-oriented architecture (SOA) one is less concerned with latency in data processing. Clearly, there are investigations of service-level agreements (SLA) and quality of service (QoS) to guarantee service delivery. Research around non-functional properties and service-level agreements for serviceoriented computing has reached a level of maturity. There are approaches for describing properties, managing SLAs and even for selecting and composing services based on NFPs. Beyond these classical topics SOA inspired extensions are enabling new and creative domains like the Internet of Things, real-time business or real-time Web. These new domains impose new requirements on SOA, such as a huge data volume, mediation between various data structures and a large number of sources that need to be procured, processed and provided. Questions like how to pick the right service out of tens of thousands of services if we talk about sensor networks or how to provide results with almost near zero-latency describe actual questions and challenges we are currently facing. Therefore, we have to look into new ways for processing data, converting and composing data coming from various sources and for enabling an easy and lightweight way to impose it on various sets of devices.

In a SOA-based system the applications are organized in a manner such that interoperable services can be used from different domains. In a process industry context, different domains can refer to, for example, process instrumentation and monitoring, execution of process control, data acquisition, etc. Large process industry systems are a complex and potentially very large sets of multi-disciplinary, heterogeneous, networked distributed systems. Current industrial process control systems are typically vendor specific; in addition the different domains are associated with different layers, different standards and different technologies. In the paper the authors report about the investigations and assessments performed to find answers for four major critical questions that arise as key when technologies have to be selected and used in a true Service Oriented Architecture (SOA) based distributed large scale Process Monitoring and Control system: (1) Real-time SOA (what are the limits of bringing SOA into high performance control loops?); (2) Management of large scale industrial distributed control systems (is it feasible to manage up to tens of thousands of service-oriented devices?); (3) Distributed event-based systems are asynchronous (what are the limits compared to traditional periodic scanning systems?) and (4) Service specification (which semantics are the most suitable for specifying process control and monitoring services?).

The last years we are witnessing of rapid advances in the industrial automation domain, mainly driven by business needs towards agility and supported by new disruptive technologies. Future factories will rely on multi-system interactions and collaborative cross-layer management and automation approaches. Such a factory, configured and managed from architectural and behavioural viewpoints, under the service-oriented architecture (SOA) paradigm is virtualized by services exposed by its key components (both HW and SW). One of the main results of this virtualization is that the factory is transformed into a “cloud of services”, where dynamic resource allocation and interactions take place. This paper presents a view on such architecture, its specification, the main motivation and considerations, as well as the preliminary services it may need to support.

The cloud concept and its implementations are gaining in importance for systems that connect evermore new devices which in turn require communication with each other. In scenarios where we can find large numbers of data providers on one side and data consumers on the other side, such as in the Internet of Things, large scale sensor networks, machine to machine communication or even in social media, one emerging requirement is to process, procure, and provide information efficiently and with almost zero latency. This work is introducing new concepts to describe the flow of data to and from sources to cloud services in a formal way by limiting information flow with filtering concepts and combining data processing techniques adopted from complex event processing.

Classic request-response Service-oriented architecture (SOA) has reached a level of maturity where SOA inspired extensions are enabling new and creative domains like the Internet of Things, real-time business or real-time Web. These new domains impose new requirements on SOA, such as a huge data volume, mediation between various data structures and a large number of sources that need to be procured, processed and provided with almost zero latency. Service selection is one of the areas where decisions have to be made based on consumer requests and service offerings. Processing this data requires typical SOA behavior combined with more elaborate approaches to process large amounts of data with near-zero latency. The approach presented in this paper combines pub-sub approaches for processing service offerings and mediations with classical request-response SOA approaches for consumer requests facilitated by Complex Event Processing (CEP). This paper presents a novel approach for subscribing to dynamic service properties and receiving up-to-date information in real-time. Therefore, we are able to select services with near-zero latency since there is no need to pull for property values anymore. The paper shows how to map requests to streaming data, how to process and answer complex requests with low latency and how to enable real-time service selection.

The Handbook of Research on Non-Functional Properties for Service-Oriented Systems: Future Directions unites different approaches and methods used to describe, map, and use non-functional properties and service level agreements. This handbook, which will be useful for both industry and academia, provides an overview of existing research and also sets clear directions for future work.